ConductorOne

No entries match these filters. Clear filters

1. October 30, 2026 (2026-10-31) Oct 31, 2026 · issue 127

   ConductorOne adds MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server discovery, Slack /c1 shortcut, IPv6IPv6The current version of the Internet Protocol, standardized by the IETF, using 128-bit addresses to replace the exhausted IPv4 space. Cyber tools target it because modern networks run dual-stack or IPv6-only. allowlists, and campaign exclusions alongside six breaking API and permission changes⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ HOW TO FIND IT
   Route SOX-tagged access requests through a dedicated approval step using the new compliance framework condition in a policy.
   📍In the Policy editor, add an approver step, open its conditions, select Compliance framework as the attribute, choose SOX, and assign the dedicated SOX approver group.
   Quickly open an access request in Slack without switching to the browser.
   copy

   $ /c1

   • Access review campaign filtering has been improved: campaigns can be filtered on resource attributes and entitlement annotations using custom selectors, combinable with app, resource type, risk level, and compliance framework filters, and grant usage filters can now be set to Used in the last [timeframe] or Unused in the last [timeframe] when building campaigns.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Exclude specific entitlements or everything matching a custom selector from campaigns and templates.
   • C1 groups gain new capabilities: review C1 group memberships — including automated groups — inside access review campaigns, and create a C1 group from an existing group or entitlement, with C1 proposing a membership rule, inclusion and exclusion lists, and flagging anyone who cannot be adopted.
   • MCP gateway visibility and error reporting have been improved: AI agents connected through the MCP gateway now see every MCP server available to them, with access status and toolsets listed per server, and MCP gateway error responses now report whether a rejected token expired, was revoked, or is malformed, returning error code -32010 when a server credential requires reauthorization.
   • Admins can now assign automatic tool access per MCP server by choosing a preset (Read tools, All tools) or creating a named toolset, and can connect external MCP servers manually by choosing Configure manually to supply custom OAuth client credentials or use a public client.

   +5 more show less

   • Start an access request from anywhere in C1 using the new sidebar New request button, which opens a right-side panel without leaving the current view.
   • Route approvals by compliance framework or app annotation — policies, approver steps, entitlement routing rules, account provisioning, and forms can now check compliance frameworks and app annotations.
   • Type /c1 in Slack to open the access request form directly, the same as /c1 request.
   • Get AI-suggested owners for unowned non-human identitynon-human identityA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. findings via Remediate with AI, which shows up to three candidates with confidence levels.
   • Turn on Use resource descriptions in an app's Entitlement management settings to show source-system descriptions instead of connector-generated text.
   • Access review campaign filtering has been improved: campaigns can be filtered on resource attributes and entitlement annotations using custom selectors, combinable with app, resource type, risk level, and compliance framework filters, and grant usage filters can now be set to Used in the last [timeframe] or Unused in the last [timeframe] when building campaigns.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Exclude specific entitlements or everything matching a custom selector from campaigns and templates.
   • C1 groups gain new capabilities: review C1 group memberships — including automated groups — inside access review campaigns, and create a C1 group from an existing group or entitlement, with C1 proposing a membership rule, inclusion and exclusion lists, and flagging anyone who cannot be adopted.
   • Admins can now assign automatic tool access per MCP server by choosing a preset (Read tools, All tools) or creating a named toolset, and can connect external MCP servers manually by choosing Configure manually to supply custom OAuth client credentials or use a public client.
   • Type /c1 in Slack to open the access request form directly, the same as /c1 request.

   +1 more show less

   • Turn on Use resource descriptions in an app's Entitlement management settings to show source-system descriptions instead of connector-generated text.

   4 more in Everything

   └──▷ BREAKING ON UPGRADE

   • !Failed client credentials token requests now always return invalid_client instead of a specific reason — scripts or SIEMsSIEMsSecurity Information and Event Management platforms that collect, normalize, and correlate log data across an infrastructure so cyber tools can detect threats and trigger alerts from a central point. matching on invalid_grant to detect deleted, disabled, or expired credentials must match invalid_client instead.
   • !Reading, updating, or creating a credential for a deleted service principal now returns not found instead of succeeding — update any scripts or Terraform that relied on the old behavior.
   • !Updating a personal API key's roles or allowed source IPs through the API now returns an error for fields that cannot be changed, instead of silently ignoring them.
   • !System log entries for failed token requests now record the real failure reason, and MCP tool calls refused by authorization are recorded as denied instead of a generic failure with code 500 — update SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. rules or alerts that matched the old status details.
   • !The email capabilities gRPCgRPCAn open-source remote procedure call framework from Google that uses HTTP/2 and Protocol Buffers to let services call each other's functions across a network, giving tools low-latency, strongly typed inter-process communication. method on the provider service is removed — clients must move to the capabilities service (GET /api/v1/settings/email-capabilities is unchanged).
   • !Application Administrators who do not own an app can no longer change an entitlement's owners, automations, or manual membership unless they are also Super Administrators; Basic Users can no longer list or download app population reports; only Super Administrators and Read-Only Administrators can read email provider settings and email activity.
   • !Policies are now limited to 16 conditional rules — existing policies above the limit remain editable but cannot grow beyond their current size.
2. Role mining suggestions (2026-10-29) Oct 29, 2026 · issue 125

   ConductorOne adds role mining suggestions that surface groups worth analyzing and automate membership enrollment for new roles.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ HOW TO FIND IT
   When onboarding a new department, use a role mining suggestion to quickly create a role that auto-enrolls future hires.
   📍1. Go to Role Mining in the C1 console. 2. Review the suggestions list — each entry shows group size, shared entitlements, and confidence score. 3. Pick a suggestion and select Open analysis — the group's filters are pre-filled. 4. Filter by user attribute (e.g. department = Engineering) or by specific app access to refine membership. 5. Create the role, then enable membership automation so new hires matching the filters receive access automatically.
   • ConductorOne now surfaces role mining suggestions showing each group's size, shared entitlements, and confidence score based on attributes such as department, job title, or manager, so analysts have a starting point instead of a blank filter bar.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Selecting a suggestion opens the analysis page with the group's filters pre-filled, letting reviewers filter by user attribute or by access held and limit the scope to specific apps.
   • Turn on membership automation for a role so users who match the group's filters — including new hires — receive the role's access automatically.
   • Any earlier role mining analysis can be resumed from the start page.

   4 details in Everything

3. changelog Oct 29, 2026 · issue 124

   ConductorOne adds a user-scoped review type, three access profile types, and a new Browse page for requestable access⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ HOW TO FIND IT
   Convert a new access profile to a Role profile so that access is granted automatically and all-or-nothing based on membership, rather than being self-service requestable.
   📍In ConductorOne, go to Access Profiles › New Profile, when prompted select the Role type, define the membership condition, and save; membership can then be granted, revoked, or requested like any other entitlement.
   • Start access review campaigns scoped to specific people — by user list, selection criteria, or CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expression — using the new By users review scope type, which also works on campaign templates to trigger reviews when a user changes roles or leaves.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Access profiles now come in three types: Catalog (self-service requestable), Role (automatic all-or-nothing grants with explicit membership), and Hybrid (the previous behavior — requestable or condition-based); all existing profiles become Hybrid automatically.
   • A new Browse page consolidates everything a user can request in one place, with a dedicated AI tools tab, while the Requests page now holds only task queues.

   3 details in Everything

4. announced Oct 29, 2026 · issue 127

   ConductorOne adds Role Mining Suggestions to identify groups worth turning into access profiles with a confidence score⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢
   └──▷ HOW TO FIND IT
   When onboarding a new team, use Role Mining Suggestions to find shared access patterns and create an access profile without manually auditing entitlements.
   📍1. Open the Role Mining page in ConductorOne. 2. Select the refresh icon to populate the suggestions list. 3. Review suggestions grouped by department, job title, or manager's team, and check each group's confidence score, user count, and shared entitlements. 4. Select a suggestion (e.g., Support Specialist at 85% confidence), then select Start › Create role to build the access profile.
   • ConductorOne now shows Role Mining Suggestions on the Role Mining page — select the refresh icon to populate a list of groups worth analyzing, ranked by confidence score and grouped by department, job title, or manager's team.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Each suggestion in the suggestions list displays user count, shared entitlements, and a confidence score so admins know which groups to prioritize before building an access profile.

   2 details in Everything

5. Campaign scope by users (2026-10-23) Oct 23, 2026 · issue 123

   ConductorOne adds a By users review scope so campaigns start from people, not entitlements.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ HOW TO FIND IT
   Trigger an access review automatically when a user changes roles or leaves, using a campaign template with By users scope paired with an automation.
   📍In the ConductorOne console, go to Access Reviews › Campaign Templates, open or create a template, set the Scope type to By users, configure user criteria or a CEL expression to match the departing/role-changing user, then go to Automations and create or edit an automation that starts this template when a user's role changes or their account is flagged for offboarding.
   • Adds a new By users review scope type to access review campaigns, letting reviewers target specific users, user criteria, or a CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expression instead of selecting individual apps and entitlements.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Supports excluding entitlements and narrowing which accounts and grants are in scope within a By users campaign.
   • Extends By users scope to campaign templates, enabling automations to trigger a review of a user's current access on role change or departure.
   • Adds a new By users review scope type to access review campaigns, letting reviewers target specific users, user criteria, or a CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expression instead of selecting individual apps and entitlements.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   2 more in Everything

6. docs update Oct 22, 2026 · issue 116

   ConductorOne: Adds a new Browse page consolidating all requestable access in one place, including a tab for AI tools.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ HOW TO FIND IT
   Request access to an AI tool for yourself or on behalf of a team member using the new Browse page.
   📍In the ConductorOne portal, go to Browse › AI tools, find the tool you need, and click Request. To request on behalf of someone else, use the Manage for option.
   • Bookmarks and links to old tabs redirect automatically to their new locations.

   1 more in Everything

7. Browse page (2026-10-21) Oct 21, 2026 · issue 116

   ConductorOne adds a dedicated Browse page with tabs for Apps, Profiles, AI tools, and Actions — centralising all requestable access in one place.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ HOW TO FIND IT
   Locate and request access to an AI tool without switching between tabs.
   📍In the ConductorOne console, navigate to Browse › AI tools, find the tool you need, and submit an access request from that page.
   • Moves app accounts to My resources > Accounts, out of the Requests page.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Refocuses the Requests page to task queues only: Assigned to me, My requests, and Temporary, opening by default on Assigned to me.
   • Managers can still browse on behalf of another user via the existing Manage view.
   • Moves app accounts to My resources > Accounts, out of the Requests page.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   2 more in Everything

   └──▷ BREAKING ON UPGRADE

   • !Requestable access (previously on the Requests page) has moved to the new Browse page; existing bookmarks and links to the old Requests tabs are redirected automatically.
8. Access profile types (2026-10-20) Oct 20, 2026 · issue 117

   ConductorOne access profiles now come in three distinct types: Catalog, Role, and Hybrid, each with purpose-built access behaviour.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ HOW TO FIND IT
   Create a Role profile to automatically provision all entitlements for a job function — no piecemeal access, membership can be requested or revoked like any other entitlement.
   📍In the C1 console, go to Access Profiles › Create Profile, select Role as the type, define a membership condition (e.g. department = Engineering), and assign the entitlements the role should grant.
   Convert an existing Hybrid profile to a Catalog profile so employees can self-service request only the specific entitlements they need.
   📍In the C1 console, open the Access Profile's details page, click Edit, change the type from Hybrid to Catalog, and save.
   • Introduces Catalog access profile type: a collection of entitlements users can self-service request from the catalog.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Introduces Role access profile type: grants entitlements automatically based on a membership condition, with membership that can be granted, revoked, and requested like other entitlements — access is all-or-nothing with no piecemeal grants.
   • Introduces Hybrid access profile type: the original C1 behaviour — entitlements requestable or auto-granted on a condition — now explicitly named; all existing profiles migrate to this type automatically.
   • Prompts users to choose a profile type on creation and allows changing the type later from the profile's details page.
   • Introduces Catalog access profile type: a collection of entitlements users can self-service request from the catalog.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Introduces Role access profile type: grants entitlements automatically based on a membership condition, with membership that can be granted, revoked, and requested like other entitlements — access is all-or-nothing with no piecemeal grants.
   • Introduces Hybrid access profile type: the original C1 behaviour — entitlements requestable or auto-granted on a condition — now explicitly named; all existing profiles migrate to this type automatically.

   1 more in Everything

9. changelog Oct 20, 2026 · issue 114

   ConductorOne adds AI-driven finding remediation, custom campaign selectors, local policies, Slack-native C1AI, and a new secure secrets-sharing feature.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

   ⤢ ⤢ ⤢ ⤢
   └──▷ TRY IT
   Safely delete only manually managed app resources via API — connector-synced resources will now return an error instead of being silently removed.
   copy

   $ curl -X DELETE 'https://<tenant>.conductor.one/api/v1/apps/<app_id>/resource_types/<resource_type_id>/resources/<id>' \
     -H 'Authorization: Bearer <token>'

   Assign C1AI as an automated evaluator step in a review policy so the agent screens every request that reaches it before a human approver is involved.
   📍In the C1 console, go to Policies › select or create a review policy › Add step › Assignee: C1AI › optionally enter custom instructions › Save.
   • Adds DELETE /api/v1/apps/{app_id}/resource_types/{resource_type_id}/resources/{id} now returns an error for connector-synced resources and exposes a read-only isManuallyManaged field on app resources returned by the API.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Adds batch Add template entitlements and Remove template entitlements API calls, each accepting up to 200 entitlements, so campaign template selections can be built incrementally without overwriting earlier choices.
   • Adds a Remediate with AI button on findings that opens C1AI with the finding's type and target pre-filled and follows a built-in remediation playbook for unowned service accounts, non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls., shadow apps, and connector sync failures, plus a new Findings tab on application detail pages listing open and in-progress findings with target, assignee, and filters for severity, type, and assignee.
   • Adds C1AI as an assignable workflow step in request and review policies, letting the agent evaluate each request using C1's built-in guidance plus custom instructions.
   • Adds secure secrets-sharing directly from C1: content is encrypted in-browser before upload, recipients authenticate via SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. or a one-time magic link, and administrators can view metadata and audit logs for all secrets across the tenant.

   +5 more show less

   • Adds C1AI assistant natively in Slack with rich interactive responses — buttons, forms, and pickers — so users can complete tasks without leaving the conversation.
   • Adds custom selectors for campaign scoping that filter by app, resource type, risk level, and compliance framework, automatically pulling in newly matching entitlements when a campaign is prepared.
   • Adds shadow apps as triageable findings, enabling assign, snooze, suppress, accept risk, reopen, and Remediate with AI actions from the Shadow apps page.
   • Adds local policies that can be created directly from the policy menu on an app or entitlement and remain bound to that single app or entitlement, outside the global policy library.
   • Adds optional character-limit setting on Text and Text area fields in request forms, preventing submission of values that exceed the limit.
   • Adds DELETE /api/v1/apps/{app_id}/resource_types/{resource_type_id}/resources/{id} now returns an error for connector-synced resources and exposes a read-only isManuallyManaged field on app resources returned by the API.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
   • Adds batch Add template entitlements and Remove template entitlements API calls, each accepting up to 200 entitlements, so campaign template selections can be built incrementally without overwriting earlier choices.
   • Adds a Remediate with AI button on findings that opens C1AI with the finding's type and target pre-filled and follows a built-in remediation playbook for unowned service accounts, non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls., shadow apps, and connector sync failures, plus a new Findings tab on application detail pages listing open and in-progress findings with target, assignee, and filters for severity, type, and assignee.
   • Adds C1AI as an assignable workflow step in request and review policies, letting the agent evaluate each request using C1's built-in guidance plus custom instructions.
   • Adds secure secrets-sharing directly from C1: content is encrypted in-browser before upload, recipients authenticate via SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. or a one-time magic link, and administrators can view metadata and audit logs for all secrets across the tenant.

   +3 more show less

   • Adds C1AI assistant natively in Slack with rich interactive responses — buttons, forms, and pickers — so users can complete tasks without leaving the conversation.
   • Adds custom selectors for campaign scoping that filter by app, resource type, risk level, and compliance framework, automatically pulling in newly matching entitlements when a campaign is prepared.
   • Adds local policies that can be created directly from the policy menu on an app or entitlement and remain bound to that single app or entitlement, outside the global policy library.

   2 more in Everything

   └──▷ BREAKING ON UPGRADE

   • !DELETE /api/v1/apps/{app_id}/resource_types/{resource_type_id}/resources/{id} now returns an error when called on connector-synced resources; previously it deleted them and they reappeared on the next sync. Any automation that depended on this deletion succeeding must be updated.
   • !Access request links (including older auto-submit links from Copilot or Jira) no longer submit requests automatically — they now only prefillprefillThe first phase of inference, where the model reads the whole prompt in parallel before producing its first token. It drives time-to-first-token; decode is the token-by-token phase after it. the form. Jira templates, runbooks, or automations that expected a link to create a request must be updated to expect an explicit user submit.
10. docs update Oct 17, 2026 · issue 111

    ConductorOne adds local policies — scope-limited policies bound to a single app or entitlement, invisible in the global policy library.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ HOW TO FIND IT
    Apply a one-off approval workflow to a single sensitive entitlement without polluting the global policy library.
    📍1. Open the entitlement that needs a unique policy. 2. Go to the Request policy setting. 3. Open the policy menu and click Create new local policy. 4. Configure the policy steps and save — it is bound to that entitlement immediately. 5. To edit later, return to the same Request policy setting and click Edit local policy.
    • Adds local policies that bind to a single app or entitlement setting (request policy, revoke policy, or preferred review policy) and never appear in the global policy library — creatable via Create new local policy from the policy menu on any app or entitlement setting.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Local policies cannot be reused, their scope cannot be changed after creation, and only Super Administrators can create or manage them.

    2 details in Everything

11. docs update Oct 17, 2026 · issue 111

    ConductorOne adds AI-assisted remediation and a per-application findings tab to the admin findings workflow.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ HOW TO FIND IT
    Get AI-guided fix steps for a specific finding without leaving the ConductorOne console.
    📍In the Findings list, open a finding's row menu and select Remediate with AI — C1AI opens with the finding's details already filled in.
    Review all open findings scoped to a single application, filtered by severity or assignee, during an access review.
    📍Navigate to the application's details page, select the Findings tab, then use the severity, type, or assignee filters to narrow the list.
    • Adds a Remediate with AI action (available from a finding's row menu and the finding details drawer) that opens C1AI with the finding's details pre-filled for step-by-step remediation guidance.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    1 detail in Everything

12. docs update Oct 17, 2026 · issue 111

    ConductorOne now includes unlinked accounts in high-risk role grant queries and CSV exports.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    • Includes accounts not linked to an identity user in high-risk role grant results for permanent grants, with blank User and email cells.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Includes accounts not linked to an identity user in high-risk role grant results for time-limited grants, with blank User and email cells.
    • Includes unlinked accounts with high-risk role grants in CSV exports, with blank User and email cells for those rows.
    • Includes accounts not linked to an identity user in high-risk role grant results for permanent grants, with blank User and email cells.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    2 more in Everything

13. docs update Oct 17, 2026 · issue 111

    ConductorOne: Adds the C1AI assistant natively in Slack, with buttons, forms, and pickers so users can complete access tasks without leaving Slack.

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Verify that deleting a connector-synced resource now surfaces an error instead of silently removing it.
    copy

    $ curl -X DELETE 'https://<tenant>.conductor.one/api/v1/apps/<app_id>/resource_types/<resource_type_id>/resources/<id>' \
      -H 'Authorization: Bearer <token>'

    Enable C1AI as a workflow step in an access review policy so the agent automatically evaluates each request that reaches it.
    📍In the C1 console, open a request or review policy, add a new workflow step, and select C1AI as the assignee. Optionally enter custom instructions the agent should follow when evaluating requests at that step.
    • App catalog search now matches entitlement descriptions, so users can find access by describing what it does rather than knowing the exact name.

    1 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !DELETE /api/v1/apps/{app_id}/resource_types/{resource_type_id}/resources/{id} now returns an error when called on connector-synced resources instead of deleting them — callers that relied on deleting synced resources will receive errors and must update their logic.
    • !Access request links (including one-click links from Copilot or Jira) no longer auto-submit requests — they now only prefillprefillThe first phase of inference, where the model reads the whole prompt in parallel before producing its first token. It drives time-to-first-token; decode is the token-by-token phase after it. the form. Jira templates, runbooks, or automations that expected a link to create a request automatically must be updated to require an explicit user submit.
14. docs update Oct 17, 2026 · issue 111

    ConductorOne adds a C1AI assistant to the product.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Access the C1AI assistant directly within ConductorOne for guided setup and operational help.
    📍In ConductorOne, locate and open the C1AI assistant from the product interface.
15. docs update Oct 17, 2026 · issue 111

    ConductorOne lets users discover and request access actions directly from the C1AI assistant in Slack, without opening the Actions catalog.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ HOW TO FIND IT
    When you need to request access without leaving Slack, ask the C1AI assistant directly instead of opening the Actions catalog.
    📍In the C1 app for Slack, open a conversation with the C1AI assistant and ask: 'What actions are available to me?' then follow the prompts to submit a request.
    • Adds the ability to ask the C1AI assistant in the C1 app for Slack what actions are available and request them directly, without navigating to the Actions catalog.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    1 detail in Everything

16. October 16, 2026 (2026-10-17) Oct 17, 2026 · issue 111

    ConductorOne adds AI-guided finding remediation, custom campaign scoping, local approval policies, and batch API for campaign templates.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Start AI-guided remediation for a finding such as an unowned service account without manually looking up remediation steps.
    📍In the Findings list, open a finding (e.g. an unowned service account), then click Remediate with AI. C1AI opens with the finding type and target pre-filled and walks through the built-in remediation playbook.
    • Adds Add template entitlements and Remove template entitlements API calls, each accepting up to 200 entitlements per call, to build or trim campaign template selections in batches without overwriting earlier choices, and introduces custom selectors for access review campaign scoping — filtering by app, resource type, risk level, and compliance framework — so new matching entitlements are pulled in automatically when a campaign is prepared.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • The DELETE /api/v1/apps/{app_id}/resource_types/{resource_type_id}/resources/{id} endpoint now returns an error for connector-synced resources instead of silently deleting them, and app resources returned by the API now include a read-only isManuallyManaged field.
    • Adds Remediate with AI on findings, opening C1AI with the finding's type and target pre-filled and following built-in playbooks for unowned service accounts, non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls., shadow apps, and connector sync failures.
    • Shadow apps are now surfaced as findings, enabling assign, snooze, suppress, accept-risk, reopen, and Remediate with AI actions on discovered apps.
    • Adds a Findings tab to application detail pages, listing open and in-progress findings with target, assignee, and filters for severity, type, and assignee.

    +5 more show less

    • Enables creation of local approval policies scoped to a single app or entitlement directly from the policy menu, keeping them out of the global policy library.
    • Adds optional character-limit settings to Text and Text area fields in request forms, preventing submission of over-length answers.
    • The validate-scope report's User grants sheet now includes the review policy, first approval step type, and expected reviewers for each review task.
    • Approvers now see custom form responses, including clickable links such as Jira tickets, directly in Slack and Microsoft Teams notifications.
    • Reviewers can open entitlement or resource details in a drawer from within an access review table without leaving the review.
    • Adds Add template entitlements and Remove template entitlements API calls, each accepting up to 200 entitlements per call, to build or trim campaign template selections in batches without overwriting earlier choices, and introduces custom selectors for access review campaign scoping — filtering by app, resource type, risk level, and compliance framework — so new matching entitlements are pulled in automatically when a campaign is prepared.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds Remediate with AI on findings, opening C1AI with the finding's type and target pre-filled and following built-in playbooks for unowned service accounts, non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls., shadow apps, and connector sync failures.
    • Adds a Findings tab to application detail pages, listing open and in-progress findings with target, assignee, and filters for severity, type, and assignee.
    • Enables creation of local approval policies scoped to a single app or entitlement directly from the policy menu, keeping them out of the global policy library.

    6 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !DELETE /api/v1/apps/{app_id}/resource_types/{resource_type_id}/resources/{id} now returns an error when called on connector-synced resources; previously it deleted them and they reappeared on the next sync.
    • !One-click access request links (such as those from Copilot or Jira) no longer submit requests automatically — they now only prefillprefillThe first phase of inference, where the model reads the whole prompt in parallel before producing its first token. It drives time-to-first-token; decode is the token-by-token phase after it. the form, requiring an explicit user submit. Jira templates, runbooks, or automations that expected auto-submit behaviour must be updated.
17. announced Oct 14, 2026 · issue 109

    ConductorOne adds AWS IAMAWS IAMAWS Identity and Access Management is Amazon's service for controlling who and what can call AWS APIs, by attaching policies that allow or deny actions on resources. Cyber tools use it to scope credentials to least privilege. Account Access Manager integration for JITJITA runtime compilation strategy where code is translated from an intermediate form to native machine instructions at execution time, giving tools the speed of compiled code without sacrificing dynamic flexibility. provisioning, access reviews, and risk detection.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢
    • Adds a new AWS IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. Account Access Manager integration that syncs IAM roles and account assignments into ConductorOne alongside existing IAM Identity Center data.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Enables bidirectional just-in-time (JIT) provisioning so employees can request time-bound access to an AWS account or IAM role through ConductorOne, with automatic revocation on expiration.
    • Surfaces AWS IAM Account Access Manager-assigned roles that are unowned, orphaned, or otherwise risky through ConductorOne's detection engine.
    • Flows AWS IAM Account Access Manager identities and role assignments into existing ConductorOne access review campaigns, so reviewers can certify AWS role assignments without a separate process.
    • Includes AWS IAM Account Access Manager-synced roles in ConductorOne's identity graph alongside SaaSSaaSSoftware delivered over the internet and run on the provider's infrastructure, so the buyer pays a subscription instead of installing or operating anything locally., other cloud platforms, and on-premises infrastructure access.
    • Adds a new AWS IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. Account Access Manager integration that syncs IAM roles and account assignments into ConductorOne alongside existing IAM Identity Center data.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    4 more in Everything

18. docs update Oct 6, 2026 · issue 096

    ConductorOne query results can now be downloaded as CSV, with selectable columns including AccountUserName and AccountDisplayName.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ HOW TO FIND IT
    Download query results as a CSV with selected columns, then join them with the accounts export on the shared username key.
    📍In the query results view, click Generate CSV, select desired columns (including AccountUserName and AccountDisplayName) in the Download as CSV drawer, then generate and download the file. Join with the accounts export on the AccountUserName column.
    • Adds Generate CSV button above query results to download them, with a Download as CSV drawer for choosing which columns to include.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Exports AccountUserName (the account's username, not display name) and a new AccountDisplayName column in the query CSV, inserted before AccountUserName, so the export can be joined with the accounts export on AccountUserName.

    2 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !The query CSV now inserts an AccountDisplayName column before AccountUserName; code that reads columns by position rather than header name will need to account for this one-column shift.
19. docs update Oct 6, 2026 · issue 096

    ConductorOne: Allows campaign templates to be configured with a specific timezone and start time instead of always beginning at 8am Pacific.

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    List scope-role bindings for an access profile catalog to audit which roles are bound to which scopes.
    copy

    $ curl -s -H 'Authorization: Bearer <token>' 'https://<tenant>.conductor.one/api/v1/catalogs/<catalog_id>/scope_role_bindings'

    └──▷ BREAKING ON UPGRADE

    • !In the Grants and Standing Privileges CSV exports, AccountUserName now contains the account username (not display name) and a new AccountDisplayName column is inserted before it — consumers reading either export by column position rather than header name must account for this one-column shift.
20. October 2, 2026 (2026-10-05) Oct 5, 2026 · issue 096

    ConductorOne adds shadow app findings, public API endpoints for attribute mappings and scope-role bindings, and c1i v0.8.0 with seven new command groups.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Fetch all scope-role bindings for an access profile catalog to audit what roles are bound to which scopes.
    copy

    $ curl -H 'Authorization: Bearer <token>' 'https://your-tenant.c1.ai/api/v1/catalogs/<catalog_id>/scope_role_bindings'

    • Adds GET /api/v1/catalogs/{catalog_id}/scope_role_bindings to the public API to read access profile scope-role bindings.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds full CRUDCRUDThe four basic database operations — Create, Read, Update, Delete — used as shorthand for any system that manages persistent data records, giving cyber tools a standard model for interacting with stored information. + list for user attribute mappings to the public API, including Terraform generation metadata.
    • Adds mcp classifiers command group to c1i v0.8.0 for MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. gateway AI governance.
    • Adds access-reviews command group to c1i v0.8.0 for managing campaigns and reports.
    • Adds findings command group to c1i v0.8.0 for public security findings and their governance rules.

    +9 more show less

    • Adds role-mining command group to c1i v0.8.0 covering analysis through access-profile creation.
    • Adds entitlements proxy-bindings command group to c1i v0.8.0.
    • Adds service-principals (alias sp) command group to c1i v0.8.0 for managing tenant-owned non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. end to end.
    • Extends access-profiles in c1i v0.8.0 with full request-catalog lifecycle management — updating or soft-deleting a profile and managing its requestable and visibility entitlements and bundle automation.
    • Introduces shadow app findings, surfacing unsanctioned apps and cloud services employees sign into with company credentials; Application Administrators can view and search all findings tenant-wide and can turn detection off.
    • Adds GetDelegates and HasDelegate CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. functions so approval policies can route requests to a user's delegate or fall back to the skip-level manager when no delegate is set.
    • Adds a search box on the Findings page that matches any substring in the Target or App columns, persists in the URL for shareable filtered views, and is respected by bulk select all matching actions.
    • Promotes Access Request Administrators' owned access profiles to the My Resources page with adjustable columns.
    • Adds timezone and start-time configuration to campaign templates so scheduled campaigns no longer always begin at 8 AM Pacific.
    • Adds GET /api/v1/catalogs/{catalog_id}/scope_role_bindings to the public API to read access profile scope-role bindings.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds full CRUDCRUDThe four basic database operations — Create, Read, Update, Delete — used as shorthand for any system that manages persistent data records, giving cyber tools a standard model for interacting with stored information. + list for user attribute mappings to the public API, including Terraform generation metadata.
    • Adds mcp classifiers command group to c1i v0.8.0 for MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. gateway AI governance.
    • Adds access-reviews command group to c1i v0.8.0 for managing campaigns and reports.
    • Adds findings command group to c1i v0.8.0 for public security findings and their governance rules.

    +5 more show less

    • Adds role-mining command group to c1i v0.8.0 covering analysis through access-profile creation.
    • Adds entitlements proxy-bindings command group to c1i v0.8.0.
    • Adds service-principals (alias sp) command group to c1i v0.8.0 for managing tenant-owned non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. end to end.
    • Introduces shadow app findings, surfacing unsanctioned apps and cloud services employees sign into with company credentials; Application Administrators can view and search all findings tenant-wide and can turn detection off.
    • Adds GetDelegates and HasDelegate CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. functions so approval policies can route requests to a user's delegate or fall back to the skip-level manager when no delegate is set.

    4 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !In the Grants and Standing Privileges CSV exports, the AccountUserName column now reports the account username instead of display name, and a new AccountDisplayName column was inserted before it — any consumer reading either export by column position rather than header name must account for this one-column shift.
21. docs update Oct 4, 2026 · issue 094

    ConductorOne adds MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server governance: auto-discovery, per-resource approval, toolset bundles, and full audit logging for AI tool calls.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ HOW TO FIND IT
    After connecting an MCP server, review its auto-discovered resources and approve them so C1 begins enforcing read entitlements on access.
    📍In the ConductorOne console, go to Admin › MCP Servers › [your server] › Resources. For each resource in Pending state, select it and choose Approve. C1 links a resource-specific read entitlement and begins checking it on every read.
    Grant a user access to a curated set of approved MCP tools by assigning a toolset through the standard request and approval workflow.
    📍In the ConductorOne console, go to Admin › MCP Servers › [your server] › Toolsets. Select an existing toolset (e.g., All approved tools - <server name>) or create a custom bundle, then submit an access request for the end user. Once approved, the toolset becomes available in the user's AI client.
    • Automatically discovers and classifies tools and resources exposed by MCP servers, with tools starting as Unset and resources starting as Pending until reviewed.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds per-resource approval gating: admins review, approve, or disable each static resource and URI template; C1 checks approval state and a linked read entitlement on every read.
    • Introduces MCP resource as a governed object — a static resource or URI template exposed by an MCP server, independent of toolsets and catalog items.
    • Introduces toolsets — named bundles of approved tools — with C1-maintained defaults (All approved tools and Read tools, suffixed with the server name) and admin-curated custom bundles.
    • Provisions end-user and agent tool access through the standard C1 request and approval workflow, surfacing approved tools to the user's AI client.

    +1 more show less

    • Logs every MCP tool call and resource read with identity, tool, and outcome for audit and compliance, plus policy context for enforcement-relevant decisions.

    6 details in Everything

22. docs update Oct 4, 2026 · issue 094

    ConductorOne now auto-creates provisioning links when delegated provisioning is configured between entitlements.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢
    └──▷ HOW TO FIND IT
    Link two entitlements so that granting Entitlement A automatically provisions Entitlement B, delegating provisioning without manual connector setup.
    📍In the ConductorOne console, navigate to the entitlement details page for Entitlement A › Provisioning › select Delegated › choose Entitlement B from the dropdown. C1 will create the provisioning link automatically. To review the link, open either entitlement's details page and view its provisioning links.
    • Adds delegated provisioning and deprovisioning: linking two entitlements so that granting or revoking one automatically grants or revokes the other, with C1 creating the provisioning link automatically.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Surfaces proposed provisioning-link changes in the UI whenever delegated provisioning is added or removed, showing both new links being created and existing links being removed when switching to manual or connector-based provisioning.

    2 details in Everything

23. docs update Oct 4, 2026 · issue 094

    ConductorOne adds an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server integration for Gemini Enterprise, enabling AI-driven C1 tool access via OAuth and Cloud Storage.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ TRY IT
    Retrieve the Gemini Enterprise web app configId so you can construct the correct tool-discovery URL before signing in.
    copy

    $ curl -s -H 'Authorization: Bearer $(gcloud auth print-access-token)' 'https://discoveryengine.googleapis.com/v1alpha/projects/YOUR_PROJECT_ID/locations/global/collections/default_collection/engines/c1-gemini-app/widgetConfigs/default_search_widget_config' | grep configId

    Check whether purchased Gemini Enterprise licenses landed in the correct project and location before attempting to assign them.
    copy

    $ curl -s -H 'Authorization: Bearer $(gcloud auth print-access-token)' 'https://discoveryengine.googleapis.com/v1alpha/billingAccounts/YOUR_BILLING_ACCOUNT_ID/billingAccountLicenseConfigs'

    • Adds support for hosting the OAuth client metadata document in a public Cloud Storage bucket, using the bucket URL as the C1 client ID (no secret, no rotation required).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds roles/discoveryengine.admin as the required Google Cloud IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. role for creating C1 data stores and managing MCP server connectors in Gemini Enterprise.
    • Adds a Gemini Enterprise data store connector that fetches the C1 tool list via MCP and enforces a maximum of 100 enabled actions per data store.
    • Supports enabling a recommended set of twelve C1 actions on the data store connector to allow the agent to find and run any tool permitted by the calling user's access profiles.
    • Adds GET /v1alpha/projects/YOUR_PROJECT_ID/locations/global/collections/default_collection/engines/YOUR_APP_ID/widgetConfigs/default_search_widget_config to retrieve the Gemini Enterprise web app configId needed to construct the tool-discovery sign-in URL.

    +3 more show less

    • Adds GET https://discoveryengine.googleapis.com/v1alpha/billingAccounts/YOUR_BILLING_ACCOUNT_ID/billingAccountLicenseConfigs to inspect license distribution across projects and locations.
    • Supports EU data residency tenants with <your-tenant>-mcp.c1eu.ai (MCP gateway) and <your-tenant>.c1eu.ai (auth endpoints) as alternate hostnames.
    • Adds the AI Governance Administrator role in ConductorOne as an alternative to Super Administrator for managing AI clients and assigning toolsets.

    8 details in Everything

24. announced Oct 2, 2026 · issue 093

    ConductorOne ships four new services: AppHub, OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. sign-in with OpenID AuthZENOpenID AuthZENAn emerging standard from the OpenID Foundation that defines a common API for querying authorization engines, letting cyber tools swap policy decision points without rewriting integration code. authz, credential vending with C1 Egress proxy, and C1 LLMLLMA large language model, a neural network trained on large amounts of text to recognize patterns and generate language. Applications use it for tasks such as answering, summarizing and writing code. Gateway.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢
    • Adds C1 AppHub, an internal deployment platform that connects apps to identity, authorization, credentials, network access, and model routing via C1.ai, with a skills file that teaches coding agents the company's governed patterns from the first commit.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds C1.ai as a standard OIDC provider that federates authentication to an existing identity provideridentity providerA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems., making app access a C1.ai entitlement subject to request, approval, review, certification, and revocation.
    • Adds an authorization endpoint built on OpenIDOpenIDAn open standard for decentralized authentication that lets a user prove their identity to a site via a trusted third-party provider, so tools can delegate login without managing credentials themselves. AuthZENAuthZENAn OpenID Foundation working group specification that standardises the API between policy enforcement points and policy decision points, letting cyber tools swap authorization engines without rewriting integration code. that lets apps query in real time whether a person can perform an action on a resource, with denials triggering an in-app access request workflow.
    • Adds credential vending that mints scoped, time-limited credentials at workload runtime, restricts them to approved IP ranges, delivers them through a vault, and provides a central audit inventory of every issuance, use, and revocation.
    • Adds C1 Egress, a governed egress proxy that injects real credentials into outbound requests only when policy allows the destination, blocks internal addresses and cloud metadata endpoints by default, and logs every call without recording the secret.

    +1 more show less

    • Adds C1 LLM Gateway, which routes model traffic by identity, data sensitivity, performance, and cost policy; attributes each inference call to a caller, a model, and a spend figure; and manages model access and budgets as entitlements with approval paths and end dates.

    6 details in Everything

25. Product digest: October 2, 2026 Oct 2, 2026 · release notes

    ConductorOne adds public API for attribute mappings and scope-role bindings, c1i v0.8.0 command groups, and enhanced automation/finding capabilities.

    • AccountUserName column in Grants and Standing Privileges CSV exports now reports the account’s username, with a new AccountDisplayName column added before it.
    • GetDelegates and HasDelegate CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. functions enable approval steps to route around out-of-office managers by delegating to a user's delegate or skip-level manager.
    • User attribute mappings are now available in the public API for read, create, update, delete, and list operations.
    • Access profile scope-role bindings are exposed via GET /api/v1/catalogs/{catalog_id}/scope_role_bindings in the public API.
    • c1i v0.8.0 adds new command groups: mcp classifiers, access-reviews, findings, role-mining, entitlements proxy-bindings, service-principals, and full lifecycle management for access-profiles.

    +2 more show less

    • Shadow app findings allow Application Administrators to view and search unsanctioned apps across the whole tenant.
    • Campaign templates now support configuring specific timezone and start time settings.

    7 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !A new AccountDisplayName column was added before AccountUserName in Grants and Standing Privileges CSV exports, shifting column positions.
26. announced Oct 1, 2026 · issue 093

    ConductorOne launches C1 LLMLLMA large language model, a neural network trained on large amounts of text to recognize patterns and generate language. Applications use it for tasks such as answering, summarizing and writing code. Gateway: a policy-controlled endpoint for routing AI inference across public, private, and customer-controlled model deployments.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢
    • Enables configurable routing rules based on provider, model, deployment, region, and data-handling requirements to determine eligible routes per workload.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Attributes inference usage and cost to the responsible caller — person, application, workload, or agent — with reporting broken down by app, department, person, or model, including cost per million tokens.
    • Supports managed credential paths to keep provider secrets out of application code when connecting to approved routes.

    3 details in Everything

27. announced Sep 30, 2026 · issue 093

    ConductorOne launches credential vending and C1 Egress proxy to eliminate hardcoded secrets from apps and AI agents.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢
    • Adds C1 Egress, a governed proxy that sits between an internal app or AI agent and the network, injecting the real credential onto a request only when the destination is allowed by policy.
    • Logs every outbound call with its source, destination, and policy decision without ever recording the secret itself.
    • Enables per-task credential issuance for AI agents, contractors, and CI/CD pipelines governed by the same policy engine used for employee access.
    • Supports instant revocation that takes effect on the next call rather than requiring a new deployment.

    4 details in Everything

28. announced Sep 29, 2026 · issue 089

    ConductorOne now handles OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. sign-in, AuthZENAuthZENAn OpenID Foundation working group specification that standardises the API between policy enforcement points and policy decision points, letting cyber tools swap authorization engines without rewriting integration code. real-time authorization, and governed access requests for AI-built apps, eliminating local user tables and stale permission snapshots.

    • New integrationActs as a standard OIDC provider for AI-built applications, federating authentication to the company's existing identity provideridentity providerA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems. without requiring a separate user table or login page.
    • New integrationAdds an OpenID AuthZENOpenID AuthZENAn emerging standard from the OpenID Foundation that defines a common API for querying authorization engines, letting cyber tools swap policy decision points without rewriting integration code. authorization endpoint so AI-built apps can query ConductorOne for real-time, per-request access decisions instead of relying on stale local role snapshots.
    • Surfaces access-denied events as governable access requests: employees can submit a request from within the application, owners review it with context, and time-bound access is granted or denied without a code change or redeploy.
    • Makes AI-built application access a ConductorOne entitlement, bringing it into existing request, approval, access-review, certification, and revocation workflows alongside all other governed systems.

    4 details in Everything

29. announced Sep 28, 2026 · issue 089

    ConductorOne launches C1 AppHub, a self-hosted internal app platform connecting AI-built apps to identity, authz, credentials, and model access.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢
    • Introduces C1 AppHub, a self-hosted internal application platform that wires each new app to ConductorOne identity, authorization, credential/token vending, network access, and model access without per-app custom integrations.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • AppHub apps are added to C1.ai's identity graph alongside every other connected system, each with an assigned owner, enabling inclusion in access reviews and offboarding workflows.
    • Supports a skills file that teaches coding agents the company's approved development pattern — directing them to use C1.ai for sign-in, authorization, credential and token vending, network access, and model access as the app is built.
    • Available as a public repository so teams can inspect the architecture, run it in their own environment, and adapt the integration pattern to their needs.

    4 details in Everything

30. changelog Sep 26, 2026 · issue 086

    ConductorOne: Removes GET /api/v1/tasks/{task_id}/spend-remedy-status; JS/TS clients must handle spend-control revision values as strings and handle new TYPE_SPEND_REMEDY = 7 task action type.

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Revoke a credential after deleting a shared secret — now a two-step operation since deletion no longer triggers revocation automatically.
    copy

    $ curl -X DELETE 'https://<tenant>.conductor.one/api/v1/secrets/<vault_id>' \
      -H 'Authorization: Bearer <token>'
    
    curl -X POST 'https://<tenant>.conductor.one/api/v1/app-secrets-admin/revoke' \
      -H 'Authorization: Bearer <token>' \
      -H 'Content-Type: application/json' \
      -d '{"vault_id": "<vault_id>"}'

    Automatically assign access findings to the right owner at classification time using the new Transformation rule step.
    📍In the C1 console, go to Governance › Transformation rules, open or create a rule, add a Set assignee step, choose the target user, and save — manual assignments on individual findings will still take precedence.
    • Renders entitlement request forms directly in the Slack modal when requesting or extending access through the C1 app for Slack.
    • Groups now display member count at the top of the group detail page.
    • Renders entitlement request forms directly in the Slack modal when requesting or extending access through the C1 app for Slack.

    1 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !A revoke criteria scoped only by a legacy resource-type name now fails the step instead of silently matching every app sharing that name — re-scope any automation still relying on name-only matching.
    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer also revokes its credential; use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately. Service SDK methods named Revoke are renamed to Delete.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs. JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7 value.
31. docs update Sep 26, 2026 · issue 086

    ConductorOne adds Revoke inherited access to trace and remove upstream group/role grants, not just direct entitlements.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢
    └──▷ HOW TO FIND IT
    Enable inherited-access revocation for an app so that removing an entitlement also removes the underlying group memberships or role assignments that granted it — preventing grants from being resurrected at the next sync.
    📍In the ConductorOne admin console, open the app, go to Settings, find the Revoke inherited access card, and enable the toggle. Once on, any revocation — manual, policy-driven, or from a denied access review campaign — will trace and remove upstream grants, and approvers will see a banner listing all downstream effects before confirming.
    • Adds Revoke inherited access toggle per app that traces entitlement access through direct grants, nested groups, role hierarchies, and delegated provisioning, revoking upstream sources automatically — not just the immediate entitlement.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Shows a banner to approvers listing downstream effects before a revocation is committed, preventing surprise removals.
    • Applies inherited-access revocation to all revocation triggers, including denied access review campaigns.
    • Adds Revoke inherited access toggle per app that traces entitlement access through direct grants, nested groups, role hierarchies, and delegated provisioning, revoking upstream sources automatically — not just the immediate entitlement.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Shows a banner to approvers listing downstream effects before a revocation is committed, preventing surprise removals.

    1 more in Everything

32. docs update Sep 26, 2026 · issue 086

    ConductorOne adds serverless TypeScript Functions for custom automation, external API calls, and identity workflow logic⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ HOW TO FIND IT
    Manually invoke a published function to test custom provisioning logic with a specific input payload.
    📍In the C1 web UI, go to Platform › Functions, select your function, click Run, provide input JSON, then click Run to execute and inspect the output JSON.
    • Retrieves API keys and secrets inside function code via functions.getConfig() for secure credential access at runtime.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Imports external npm packages dynamically inside functions using the npm: prefix.
    • Accesses pre-authenticated C1 APIs inside functions via the @c1/functions-sdk package.
    • Introduces serverless TypeScript Functions that run custom automation logic, integrate with external systems, and implement organization-specific identity governance workflows.
    • Triggers functions from automation workflows on user lifecycle events (created, updated, deactivated), access events (grant created, revoked), review events (completed, decisions made), and schedule triggers (daily, weekly, custom).

    +5 more show less

    • Runs functions on-demand from the web UI via Platform > Functions, with input and output JSON.
    • Invokes functions programmatically via the C1 API for administrative tasks, batch processing, or external integrations.
    • Controls which external domains functions may reach via a network allowlist.
    • Executes each function in an isolated sandboxed environment authenticated as a service principal, with role bindings determining runtime permissions.
    • Streams console.log execution logs back to the UI.

    10 details in Everything

33. docs update Sep 26, 2026 · issue 086

    ConductorOne adds Dynamic Access Control with auto-enrollment and auto-unenrollment via JMLJMLA formal specification language for Java that lets developers write machine-checkable contracts — preconditions, postconditions, and invariants — directly in source comments, enabling static verification and test generation. membership automation on access profiles⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Configure a contractor access profile to auto-enroll any user who holds a Contractor entitlement, auto-approve enrollment, and immediately revoke unjustified access when they leave — with a safeguard that pauses if more than 10% of members would be removed in one hour.
    📍1. Open the access profile and go to its JML page. 2. In the Joiner section, click Edit, select your auto-approve policy for the enrollment entitlement, choose Bypass each entitlement's approval policy, set Maximum membership duration if needed, then click Save. 3. In the Leaver section, click Edit, select your revocation policy, set post-unenrollment action to Revoke unjustified entitlements, choose Bypass each entitlement's revoke policy, then click Save. 4. In the Membership automation area, click Edit, select the Contractor entitlement under Membership automation, enable 'Grant and revoke entitlements on membership change', set Removed members threshold (%) to 10 in the Anomaly detection section, then click Save. 5. Flip Enabled to On and click Save.
    Force an immediate membership sync after adding a new entitlement to an access profile, rather than waiting up to an hour for the next scheduled run.
    📍On the access profile's JML page, click Sync now.
    Resume a paused membership automation sync after reviewing staged removals that triggered the anomaly-detection safeguard.
    📍1. On the main Access profile page, locate the access profile showing a safeguard alert banner. 2. Open the profile and click Review in the banner. 3. Inspect the list of members staged for removal. 4. If the list looks correct, click Resume sync; otherwise click Go back and adjust the Membership automation settings.
    • Adds a JML page on each access profile where admins configure joiner rules (enrollment approval policy, optional Maximum membership duration) and leaver rules (revocation policy, post-unenrollment entitlement handling).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds Membership automation configuration on the access profile JML page: select entitlements as the membership condition, define an Excluded users list, and toggle Grant and revoke entitlements on membership change to drive auto-enrollment and auto-unenrollment.
    • Adds anomaly-detection safeguard (Pause automation when anomalies are detected) that pauses any membership sync removing more than 20% of members within a single hour for profiles with 100 or more members; the cutoff is adjustable via Removed members threshold (%) in the Anomaly detection section.
    • Adds a Sync now button on the access profile's JML page to trigger an immediate membership automation sync outside the default hourly cadence.
    • Adds a pending members count and drill-through on the access profile's Members page to surface users whose enrollment request tasks are awaiting approval.

    +4 more show less

    • Adds a Review / Resume sync workflow on paused access profiles, with banner alerts in the UI and a Slack notification to C1 admins when a safeguard triggers.
    • Leaver rules support three post-unenrollment entitlement actions: leave access as-is, revoke all entitlements granted by the access profile, or revoke only unjustified entitlements (those not duplicated in another access profile the user still holds).
    • Joiner and leaver rules each support a 'bypass individual entitlement policies' option, treating approval of the enrollment or unenrollment entitlement as implicit approval for all contained entitlements.
    • Adds an Enroll users button on the access profile's Members page to manually add users who do not meet the membership automation condition.

    9 details in Everything

34. docs update Sep 26, 2026 · issue 086

    ConductorOne adds Access Profiles for grouping entitlements, role mining recommendations, and granular enrollment expiration controls⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢
    └──▷ HOW TO FIND IT
    Create a department-scoped access profile that only Engineering staff (users holding a specific entitlement) can see and request, keeping the catalog tidy for everyone else.
    📍In ConductorOne, go to Governance › Access profiles and click New profile. After naming it, open the Requests tab, click Edit in the Allow requests area, toggle on Allow requests, set Profile is visible to to Users with entitlements, and select the Engineering entitlement from the dropdown. Click Save.
    Enforce time-limited access for contractors by capping how long they can hold a profile and automatically revoking grants that aren't covered by another active enrollment when it expires.
    📍Open the access profile's JML tab. Set Maximum membership duration to your desired cap (e.g. 90 days). Set Access change behavior to Revoke unjustified entitlements. Save.
    Retroactively grant a newly added entitlement to all users already enrolled in a profile without waiting for their next renewal.
    📍On the access profile's Apps tab, locate the newly added entitlement, click the '...' (more actions) menu, select Manage provisioning, then click Request for all to create access request tasks for every enrolled user who lacks the entitlement.
    • Adds Access Profiles (Governance > Access profiles) to group apps and entitlements into requestable bundles scoped by audience — all org members or only users who hold specific entitlements.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds profile membership requests via the Profiles tab on the New request form and Requests page, letting users request all entitlements in a profile at once for onboarding scenarios.
    • Adds Maximum membership duration and Access change behavior settings on the access profile's JMLJMLA formal specification language for Java that lets developers write machine-checkable contracts — preconditions, postconditions, and invariants — directly in source comments, enabling static verification and test generation. tab to cap enrollment length and control what happens to grants when enrollment expires.
    • Adds three Access change behavior options — Leave access as-is, Revoke all entitlements, and Revoke unjustified entitlements — giving admins precise control over post-enrollment grant lifecycle.
    • Adds role mining recommendations with confidence scores (80%+ strong fit, 50–79% possible fit, 49% or lower poor fit) to suggest entitlements for access profiles based on org-wide access patterns.

    +3 more show less

    • Adds Create requests for currently enrolled users checkbox when saving entitlement changes to an access profile, automatically generating access request tasks for all enrolled users.
    • Adds Manage provisioning action on the access profile Apps tab to manually request new entitlements for current profile holders who do not yet have them, including a Request for all bulk option.
    • Enables adding an entitlement to one or more access profiles directly from the entitlement's details page via the Access profiles field in the Access requests section.

    8 details in Everything

35. docs update Sep 26, 2026 · issue 086

    ConductorOne adds secrets inventory, zero-knowledge secret sharing, workload federation, and tripwire decoy credentials.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    └──▷ HOW TO FIND IT
    Review unrotated or orphaned API keys across all connected SaaSSaaSSoftware delivered over the internet and run on the provider's infrastructure, so the buyer pays a subscription instead of installing or operating anything locally. systems before a compliance audit.
    📍In the ConductorOne console, go to Inventory › Secrets tab — review flagged secrets that have not been used recently, noting the attributed identity for each, then initiate revocation or rotation for any orphaned keys.
    • Adds a Secrets tab on the Inventory dashboard to track API tokens and service account keys across connected systems (Okta, GitHub, AWS, Snowflake, Datadog, and Google Cloud Platform), attributed to owning identities with alerts for recently-unused or orphaned keys.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds credential stores that hold provisioning-generated passwords created when C1 provisions a new account through a connector, decryptable by store owners or the account recipient.
    • Adds zero-knowledge secret sharing for one-time delivery of any secret — passwords, API keys, JSON service-account files, .env blocks — to a named recipient inside or outside the org; encryption happens in the sender's browser, C1 never sees plaintext, and access expires after a configured time or view count.
    • Adds workload federation so CI/CD pipelines and workloads can exchange a platform OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. token (GitHub ActionsGitHub ActionsGitHub's built-in CI/CD system that runs automated workflows triggered by repository events, letting cyber tools automate testing, building, and releasing without managing separate pipeline infrastructure., GitLab CI, HCP Terraform, SPIFFESPIFFEAn open standard, maintained by the CNCF, for assigning cryptographic identities to workloads in dynamic infrastructure so services can authenticate each other without passwords or static credentials., or any custom OIDC issuer) for a short-lived C1 access tokenaccess tokenA credential that authorizes a client to use a service or API. It usually has limited permissions and a lifetime; anyone holding it may be able to use those permissions. scoped to a service principal's roles — no stored client secret required.
    • Adds enterprise-managed authorization for AI agents reaching an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server, exchanging the signed-in user's identity for a short-lived, entitlement-checked token scoped to that one server.

    +1 more show less

    • Adds Decoys — tripwire credentials shaped like real human, service-principal, connector, workload-federation, or long-lived-token credentials — that raise a Critical finding automatically when used, as a high-confidence indicator of compromiseindicator of compromiseAn indicator of compromise, an observable clue that may point to malicious activity, such as a suspicious file hash or domain. It is evidence to investigate, rather than proof by itself..

    6 details in Everything

36. changelog Sep 25, 2026 · issue 085

    ConductorOne adds EU data residency, AI task diagnostics, public REST APIREST APIA web API design style where clients interact with resources over HTTP using standard methods like GET, POST, and DELETE, giving cyber tools a predictable, language-agnostic way to integrate with external services. for AI governance, and new MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. Terraform resources alongside breaking API changes.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Revoke a credential after deleting a shared secret — two calls are now required where one used to suffice.
    copy

    $ curl -X DELETE 'https://<tenant>.conductor.one/api/v1/secrets/{vault_id}' \
      -H 'Authorization: Bearer <token>'
    
    curl -X POST 'https://<tenant>.conductor.one/api/v1/app-secrets-admin/revoke' \
      -H 'Authorization: Bearer <token>' \
      -H 'Content-Type: application/json' \
      -d '{"vault_id": "<vault_id>"}'

    Save a C1AI reporting conversation so your team can revisit and refresh the same query without re-prompting.
    📍In the ConductorOne console, open C1AI and ask your reporting question, then select Save report, give it a name and description, and find it later under Platform › Reporting where you can refresh it or edit it with the Report assistant.
    • Adds POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately from secret deletion, after DELETE /api/v1/secrets/{vault_id} is scoped to delete-only.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Exposes classifiers, tool gates, classifier bindings, templates, and the agent classifier policy through the public REST API for AI governance management.
    • Adds Terraform resources for hosted and external MCP sources, covering details, authentication, and tool enablement policy via C1's Terraform provider.
    • Introduces an EU data residency instance with region-scoped login, redirect URIs, API and webhook endpoints, MCP server, and CLI configuration — selected at domain registration.
    • Adds a Set assignee step to Transformation rules so findings can be auto-assigned to a specific person (manual assignment takes precedence).

    +14 more show less

    • Adds Platform > Reporting page where saved C1AI reports can be refreshed, edited with the Report assistant, or inspected for data sources.
    • MCP access now requires an explicit toolset grant on the server — app membership or admin role alone no longer implies access.
    • Adds a Change history icon to connectors, policies, entitlement routing rules, SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. and branding settings, and AI Gateway spend controls, showing a before-and-after timeline per field.
    • Connectors can now require a justification whenever their configuration changes, captured alongside change history.
    • Super Admins can now ask C1AI to diagnose a stuck, failed, or waiting task, tracing approvals, provisioning, access-profile origin, and connector handoff.
    • Access-review revoke tasks now automatically include the reviewer's justification in ticketing system output, with new template variables available for custom templates.
    • Adds Apps and Entitlements tabs to the My resources page alongside existing groups, vaults, credential stores, and secrets.
    • Registered MCP integrations — catalog and external — now share a single guided setup flow for connection, access, tool approval, and authentication.
    • Campaigns can now include users with pending account invitations instead of silently dropping them.
    • Customer agents connected to the C1 MCP endpoint can now discover and converse with c1aw agents, with admin session management.
    • Conflict monitors now support groups of very different sizes (e.g., 90 and 11) as long as the total stays under 1024.
    • Request forms configured on entitlements now render and submit directly in the Slack modal for the C1 app for Slack.
    • Adds a Resource description column to access-review tables, showing custom descriptions or connector-synced metadata as a fallback.
    • Introduces TYPE_SPEND_REMEDY = 7 as a new task action type value in the task APIs.
    • Adds POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately from secret deletion, after DELETE /api/v1/secrets/{vault_id} is scoped to delete-only.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Exposes classifiers, tool gates, classifier bindings, templates, and the agent classifier policy through the public REST API for AI governance management.
    • Adds Terraform resources for hosted and external MCP sources, covering details, authentication, and tool enablement policy via C1's Terraform provider.
    • Introduces an EU data residency instance with region-scoped login, redirect URIs, API and webhook endpoints, MCP server, and CLI configuration — selected at domain registration.
    • Adds Platform > Reporting page where saved C1AI reports can be refreshed, edited with the Report assistant, or inspected for data sources.

    +6 more show less

    • Adds a Change history icon to connectors, policies, entitlement routing rules, SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. and branding settings, and AI Gateway spend controls, showing a before-and-after timeline per field.
    • Connectors can now require a justification whenever their configuration changes, captured alongside change history.
    • Super Admins can now ask C1AI to diagnose a stuck, failed, or waiting task, tracing approvals, provisioning, access-profile origin, and connector handoff.
    • Customer agents connected to the C1 MCP endpoint can now discover and converse with c1aw agents, with admin session management.
    • Adds a Resource description column to access-review tables, showing custom descriptions or connector-synced metadata as a fallback.
    • Introduces TYPE_SPEND_REMEDY = 7 as a new task action type value in the task APIs.

    8 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !A revoke criteria scoped only by a legacy resource-type name now fails the automation step closed instead of silently matching every app sharing that name — re-scope any automation still relying on name-only matching.
    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer also revokes its credential; use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately. Service SDKs rename Revoke to Delete to match.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs. JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7.
37. changelog Sep 24, 2026 · issue 084

    ConductorOne: Adds a single guided setup flow for creating apps and adding connectors.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Automatically route findings to the right owner in a Transformation rule instead of relying on manual assignment.
    📍In the C1 console, go to Automation › Transformation rules, open or create a rule, add a Set assignee step, and select the target user — this assignment takes precedence over any later manual change.
    • Adds searchable step and trigger-type search by name or description in the automation builder.
    • Adds faceted, searchable resource-type pickers with real counts to Grant Found, Grant Deleted, and revoke criteria.

    2 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !A revoke criteria scoped only by a legacy resource-type name now fails closed (fails the step) instead of silently matching every app sharing that name — re-scope any automation still relying on name-only matching.
    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer also revokes its credential; use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately. Service SDK Revoke method is renamed to Delete.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs. JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7.
38. changelog Sep 23, 2026 · issue 083
    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Navigate to the Change history for a connector to audit who changed its configuration and what changed.
    📍In the C1 console, go to Connectors › [select a connector] › detail page, then click the Change history icon to view the per-field before-and-after timeline.
    Save a C1AI report for recurring use by your access-review team.
    📍In the C1 console, go to Platform › Reporting, open the Report assistant, ask your reporting question, then select Save report and enter a name and description.
    └──▷ BREAKING ON UPGRADE

    • !A revoke criteria scoped only by a legacy resource-type name now fails closed (the step fails) instead of silently matching every app sharing that name — re-scope any automation still relying on name-only matching.
    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer also revokes its credential; use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately. Service SDKs rename Revoke to Delete to match.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs. JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7.
39. changelog Sep 22, 2026 · issue 082

    ConductorOne adds EU data residency, saveable AI reports, REST APIREST APIA web API design style where clients interact with resources over HTTP using standard methods like GET, POST, and DELETE, giving cyber tools a predictable, language-agnostic way to integrate with external services. for AI governance, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. Terraform support, and a POST /api/v1/app-secrets-admin/revoke credential-revoke endpoint.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Revoke a credential after deleting a shared secret — now two separate calls since deletion no longer implies revocation.
    copy

    $ curl -X DELETE 'https://<tenant>.conductor.one/api/v1/secrets/<vault_id>' \
      -H 'Authorization: Bearer <token>'
    
    curl -X POST 'https://<tenant>.conductor.one/api/v1/app-secrets-admin/revoke' \
      -H 'Authorization: Bearer <token>' \
      -H 'Content-Type: application/json' \
      -d '{"vault_id": "<vault_id>"}'

    Diagnose a stuck provisioning task as a Super Admin without touching any live state.
    📍In the C1 console, open the stuck task's detail page, then select Ask C1AI and type: 'Why is this task stuck and what should I do next?'
    Automatically assign access-review findings to the right owner using a Transformation rule instead of manual triage.
    📍In the C1 console, go to Automation › Transformation rules, open or create a rule, add a new step, select Set assignee, and specify the target user — this assignment will override any automatic assignee for matched findings.
    • Adds POST /api/v1/app-secrets-admin/revoke endpoint to queue credential revokes separately from secret deletion, replacing the previous combined behavior of DELETE /api/v1/secrets/{vault_id}.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Exposes classifiers, tool gates, classifier bindings, templates, and the agent classifier policy through the public REST API for AI governance management.
    • Introduces TYPE_SPEND_REMEDY = 7 as a new task action type; JS/TS clients must handle spend-control revision values as strings.
    • Adds Terraform resources for managing hosted and external MCP sources — including details, authentication, and tool enablement policy — as code via C1's Terraform provider.
    • Launches EU data residency instance, selectable at domain registration, affecting all tenant-specific URLs including login, redirect URIs, API and webhook endpoints, MCP server, and CLI configuration.

    +14 more show less

    • Adds Save report action and a Platform > Reporting page where saved C1AI reports can be named, described, refreshed, and edited with the Report assistant.
    • Enables Super Admins to ask C1AI to diagnose a stuck, failed, or waiting task — tracing approvals, provisioning, access-profile origin, and connector handoff.
    • Adds a Change history icon to configuration object detail pages (connectors, policies, entitlement routing rules, SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management., branding settings, AI Gateway spend controls) showing a before-and-after timeline of every change.
    • Adds optional justification requirement for connector configuration changes, captured alongside change history.
    • Adds a Set assignee step to Transformation rules for automatically assigning a finding to a specific person.
    • Adds Apps and Entitlements tabs to the My resources page, consolidating groups, vaults, credential stores, secrets, apps, and entitlements in one view.
    • Adds agent-to-agent discovery: customer agents connected to a C1 MCP endpoint can now discover and converse with c1aw agents.
    • Changes MCP access control to require an explicit toolset grant on the MCP server — app membership or admin role alone no longer implies access.
    • Unifies registered MCP integrations (catalog and external) into a single guided setup flow covering connection, access, tool approval, and authentication.
    • Renders entitlement request forms inline in the Slack modal when requesting or extending access via the C1 app for Slack.
    • Includes access-review revoke-task reviewer denial reasons in ticketing system templates automatically, with new template variables available for custom templates.
    • Expands campaign scope to include users with pending (unaccepted) account invitations.
    • Allows conflict monitors to handle groups of very different sizes (e.g., 90 and 11) as long as the combined total stays under 1,024 members.
    • Adds a Resource description column to access-review tables, showing custom descriptions or connector-synced metadata as a fallback.
    • Adds POST /api/v1/app-secrets-admin/revoke endpoint to queue credential revokes separately from secret deletion, replacing the previous combined behavior of DELETE /api/v1/secrets/{vault_id}.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Exposes classifiers, tool gates, classifier bindings, templates, and the agent classifier policy through the public REST API for AI governance management.
    • Introduces TYPE_SPEND_REMEDY = 7 as a new task action type; JS/TS clients must handle spend-control revision values as strings.
    • Adds Terraform resources for managing hosted and external MCP sources — including details, authentication, and tool enablement policy — as code via C1's Terraform provider.
    • Launches EU data residency instance, selectable at domain registration, affecting all tenant-specific URLs including login, redirect URIs, API and webhook endpoints, MCP server, and CLI configuration.

    +6 more show less

    • Adds Save report action and a Platform > Reporting page where saved C1AI reports can be named, described, refreshed, and edited with the Report assistant.
    • Enables Super Admins to ask C1AI to diagnose a stuck, failed, or waiting task — tracing approvals, provisioning, access-profile origin, and connector handoff.
    • Adds a Change history icon to configuration object detail pages (connectors, policies, entitlement routing rules, SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management., branding settings, AI Gateway spend controls) showing a before-and-after timeline of every change.
    • Adds optional justification requirement for connector configuration changes, captured alongside change history.
    • Adds a Set assignee step to Transformation rules for automatically assigning a finding to a specific person.
    • Adds agent-to-agent discovery: customer agents connected to a C1 MCP endpoint can now discover and converse with c1aw agents.

    8 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer revokes its credential — use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately. Service SDKs rename Revoke to Delete to match.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs. JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7.
    • !A revoke criteria scoped only by a legacy resource-type name now fails the step instead of silently matching every app sharing that name — re-scope any automation still relying on name-only matching.
40. docs update Sep 22, 2026 · issue 082

    ConductorOne AI assistant gains persistent saved reports and read-only stuck-task diagnosis for Super Admins.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Diagnose a task that is stuck in provisioning without taking any disruptive action — useful before escalating or retrying.
    📍In the C1AI assistant, type: 'why is task #4821 stuck?' The assistant traces the task's approval and provisioning state, access profile origin, app-account access, and connector handoff, then recommends whether to approve, deny, or retry.
    Find and re-run a previously saved access report without hunting through conversation history.
    📍Navigate to Platform › Reporting, locate the saved report by name, then select Refresh to re-run it against the latest data, or Edit with Report assistant to adjust its parameters in a conversation.
    • Adds Save report action in the AI assistant conversation to name, persist, and later refresh or edit reports without losing them between sessions.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds a Platform > Reporting page listing all saved reports, with rename and delete controls, so reports are discoverable outside conversation history.
    • Adds a Function tab on each saved report's page to inspect the underlying logic and data sources behind every result.
    • Adds read-only task diagnosis for Super Admins: ask C1AI why a specific task is stuck, failed, waiting, or in an unexpected access state (e.g. 'why is task #4821 stuck?') and receive a traced history plus a recommended next step.

    4 details in Everything

41. docs update Sep 22, 2026 · issue 082

    ConductorOne: Adds a Change history icon to connector, policy, entitlement routing rule, SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management., branding, and AI Gateway spend-control detail pages, with optional justification requirement for connector config changes.

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Revoke a credential after deleting a shared secret, using the new two-step API now that deletion and revocation are separate operations.
    copy

    $ curl -X DELETE 'https://<tenant>.conductor.one/api/v1/secrets/<vault_id>' \
      -H 'Authorization: Bearer <token>'
    
    curl -X POST 'https://<tenant>.conductor.one/api/v1/app-secrets-admin/revoke' \
      -H 'Authorization: Bearer <token>' \
      -H 'Content-Type: application/json' \
      -d '{"vault_id": "<vault_id>"}'

    Diagnose a stuck or failed task without manually combing through approval history and provisioning logs.
    📍In C1, open the stuck task's detail page, then select Ask C1AI and type: 'Why is this task stuck and what should I do next?'
    • Adds a member-count display at the top of the group detail page.
    • Accounts with outstanding pending invitations can now be included in campaign scope instead of being silently dropped.

    2 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !A revoke criteria scoped only by a legacy resource-type name now fails the automation step instead of silently matching every app sharing that name — re-scope any automation still relying on name-only matching.
    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer also revokes its credential; use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately. Service SDKs rename Revoke to Delete to match.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs. JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7 value.
42. docs update Sep 19, 2026 · issue 079

    Findings rules can now auto-assign, set severity, and add or remove annotations as findings arrive⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    • Findings rules now support automatically setting severity, adding annotations, removing annotations, or assigning every matching finding to a specific person as it arrives (a person's own manual assignment always overrides a later rule).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    1 detail in Everything

43. docs update Sep 19, 2026 · issue 079

    ConductorOne AI assistant gains read-only task diagnosis for stuck, failed, or unexpected-state tasks.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    When a provisioning task is stuck or shows unexpected access state, ask C1AI to trace its full history and get a recommended remediation step before manually acting.
    📍In the C1AI assistant, type: 'why is task #4821 stuck?' — the assistant returns its approval and provisioning state, access profile origin, app-account access, and connector handoff details, then recommends whether to approve, deny, or retry the task.
44. docs update Sep 19, 2026 · issue 079

    ConductorOne: Removes GET /api/v1/tasks/{task_id}/spend-remedy-status — migrate to the task APIs; JS/TS clients must handle spend-control revision values as strings and integrations must handle the new TYPE_SPEND_REMEDY = 7 action type.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Diagnose a stuck or failing task as a Super Admin to get C1AI's recommended next step without making changes.
    📍In the C1 console, open the stuck task, then select Ask C1AI and choose Diagnose this task to trace its approval history, provisioning log, and connector handoff.
    • Launches EU data residency instance — choose your hosting region at domain registration to scope all tenant URLs (login, redirect URIs, API, webhooks, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server, CLI) to the EU.
    • Groups now display member count at the top of the group detail page.
    • Launches EU data residency instance — choose your hosting region at domain registration to scope all tenant URLs (login, redirect URIs, API, webhooks, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server, CLI) to the EU.

    1 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !A revoke criteria scoped only by a legacy resource-type name now fails the step closed instead of silently matching every app sharing that name — re-scope any automation still relying on name-only matching.
    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer also revokes its credential — use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately; service SDKs rename Revoke to Delete to match.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs; JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7.
45. September 18, 2026 (2026-09-18) Sep 18, 2026 · issue 079

    ConductorOne adds C1AI task diagnostics, finding assignment, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. Terraform resources, and a public REST APIREST APIA web API design style where clients interact with resources over HTTP using standard methods like GET, POST, and DELETE, giving cyber tools a predictable, language-agnostic way to integrate with external services. for AI governance⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    • Adds POST /api/v1/app-secrets-admin/revoke endpoint to queue a credential revoke separately from secret deletion.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Exposes classifiers, tool gates, classifier bindings, templates, and the agent classifier policy through the public REST API for AI governance management.
    • Adds Terraform resources for hosted and external MCP sources, covering details, authentication, and tool enablement policy.
    • Adds a Set assignee step in Transformation rules to automatically assign a finding to a specific person.
    • Super Admins can now ask C1AI to diagnose a specific stuck, failed, or waiting task, tracing approval history, provisioning logs, access-profile origin, and connector handoff.

    +12 more show less

    • Adds Apps and Entitlements tabs to the My resources page, consolidating groups, vaults, credential stores, secrets, apps, and entitlements — including secrets shared with you — in one view.
    • Adds a single guided setup flow for creating apps and adding connectors, replacing the multi-page form.
    • Admins can now require a justification whenever a connector's configuration changes, with a secret-safe configuration history.
    • Customer agents connected to your C1 MCP endpoint can now discover and converse with c1aw agents, with admin-managed sessions.
    • MCP access now requires an explicit grant on a server's toolset — app membership or admin role alone no longer implies access.
    • Registered MCP integrations (catalog and external) now share a single guided setup flow for connection, access, tool approval, and authentication.
    • Access-review revoke tasks now automatically include the reviewer's denial justification, surfacing it in ticketing-system templates via new template variables.
    • Campaigns can now include users with pending (unaccepted) account invitations in their scope.
    • Conflict monitors can now handle groups of very different sizes as long as the total stays under 1,024 members.
    • Reviewers can now see a Resource description column in review tables, showing custom descriptions or connector-synced metadata.
    • Resource-type filters in Grant Found, Grant Deleted, and revoke criteria now use a faceted, searchable picker with real counts and name disambiguation.
    • Automation builders can now search available steps and trigger types by name or description.
    • Adds POST /api/v1/app-secrets-admin/revoke endpoint to queue a credential revoke separately from secret deletion.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Exposes classifiers, tool gates, classifier bindings, templates, and the agent classifier policy through the public REST API for AI governance management.
    • Adds Terraform resources for hosted and external MCP sources, covering details, authentication, and tool enablement policy.
    • Adds a Set assignee step in Transformation rules to automatically assign a finding to a specific person.
    • Super Admins can now ask C1AI to diagnose a specific stuck, failed, or waiting task, tracing approval history, provisioning logs, access-profile origin, and connector handoff.

    +2 more show less

    • Admins can now require a justification whenever a connector's configuration changes, with a secret-safe configuration history.
    • Customer agents connected to your C1 MCP endpoint can now discover and converse with c1aw agents, with admin-managed sessions.

    10 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !A revoke criteria scoped only by a legacy resource-type name now fails the step instead of silently matching every app sharing that name — re-scope any automation relying on name-only matching.
    • !DELETE /api/v1/secrets/{vault_id} now deletes only the secret and no longer revokes its credential; use the new POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately. Service SDKs rename Revoke to Delete to match.
    • !GET /api/v1/tasks/{task_id}/spend-remedy-status has been removed — migrate to the task APIs. JS/TS clients must handle spend-control revision values as strings, and integrations with a closed list of task action types must handle the new TYPE_SPEND_REMEDY = 7.
46. Product digest: September 18, 2026 Sep 18, 2026 · release notes

    C1AI report saving, task diagnosis, change history, finding assignment, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. agent-to-agent comms, Slack request forms, and API governance

    • Save report action in C1AI conversations to name and revisit analysis results later on the Platform > Reporting page
    • Diagnose a stuck task capability for Super Admins where C1AI traces approval, provisioning, and connector history without making changes
    • Change history timeline with before-and-after views for connectors, policies, entitlement routing rules, SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management., branding settings, and AI Gateway spend controls, including a new justification requirement for connector configuration changes
    • Set assignee step in Transformation rules (overriding manual assignments) to assign findings to specific people
    • MCP agent-to-agent conversation capability allowing customer agents connected to the C1 MCP endpoint to discover and hold conversations with c1aw agents

    +1 more show less

    • Explicit toolset grant requirement for MCP access, decoupling server access from app membership or admin roles

    6 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !Revoke criteria scoped only by a legacy resource-type name now fails the step instead of silently matching apps with that name
    • !DELETE /api/v1/secrets/{vault_id} no longer revokes credentials; use POST /api/v1/app-secrets-admin/revoke to queue a credential revoke separately
47. docs update Sep 15, 2026 · issue 075

    ConductorOne adds an EU data residency instance for organizations requiring data stored and processed in the European Union.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    • Adds an EU data residency instance where all tenant-specific URLs — login, redirect URIs, API and webhook endpoints, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server, and CLI configuration — are region-scoped; hosting region is selected at domain registration time.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    1 detail in Everything

48. docs update Sep 15, 2026 · issue 075

    ConductorOne: Adds an EU data residency instance at yourcompany.c1eu.ai for organizations that need customer data stored and processed in the EU, with dedicated auth, API, webhook, and MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. endpoints under c1eu.ai.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    • EU tenants use https://accounts.c1eu.ai/auth/callback as the redirect/callback URI and https://your_domain.c1eu.ai/login as the initiate login URI across SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. and OAuth integrations.
    • EU tenants sign up at accounts.c1eu.ai/accounts/signup — hosting region is a one-time choice made at signup and cannot be changed afterward.

    2 details in Everything

49. announced Sep 15, 2026 · issue 078

    ConductorOne is now available as an EU-hosted instance, with tenant data held in AWS Frankfurt.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢
    • New EU instance of C1.ai provisions tenant data in AWS Frankfurt (Ireland as disaster-recovery region), enabling European customers to meet data-residency requirements.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    1 detail in Everything

50. EU instance (2026-09-14) Sep 14, 2026 · issue 075

    ConductorOne adds an EU data residency instance with region-specific URLs for login, API, webhooks, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface., and CLI.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    When registering a new ConductorOne domain for an EU-based organization, select the EU hosting region so that all tenant URLs (login, API, webhooks, MCP server, CLI) are routed to EU infrastructure.
    📍In the ConductorOne domain registration flow, under Hosting region, select EU before completing registration. All tenant-specific URLs — including login, redirect URIs, API and webhook endpoints, MCP server, and CLI configuration — will reflect the EU instance from that point on.
    • Microsoft Teams as a notification and interaction surface is not yet supported on the EU data residency instance.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    1 detail in Everything

51. changelog Sep 14, 2026 · issue 074

    ConductorOne: Adds per-section overview dashboards across all major admin nav sections, replacing what previously lived in the single Product digest view.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Cap how many entitlements a user can bulk-request at once to reduce approval noise — for example, limit to 5 per request.
    📍In the C1 admin console, go to Governance › Forms, open Tenant-wide request settings, and set the bulk request limit to 5 (range: 1-100, default 10).
    └──▷ BREAKING ON UPGRADE

    • !Automation expressions that reference the bare workflow context, or pass it directly to a function, no longer compile — expressions must name the step and field directly. Any automation using an indirect-reference expression will fail to save on next edit until rewritten to a static path.
    • !The Owner column, filter, detail field, API field, and audit-log event names on Findings are renamed to Assignee — integrations consuming the Finding object via API, webhook, or SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. audit log must update to the new field and event names.
52. changelog Sep 13, 2026 · issue 073

    ConductorOne adds agent classifiers, four session-policy API endpoints, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. resource governance, and new system log event classes.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Enable an agent classifier in Observe mode first so you can review what calls it would have blocked before allowing it to deny anything in production.
    📍In the C1 admin console, go to AI › Classifiers, create or open a classifier, write rules over private-data sensitivity, untrusted-content exposure, and exfiltration capability, then set the enforcement mode to Observe and save.
    Raise the per-user bulk request limit for your tenant so power users can request more entitlements in one submission without multiple round-trips.
    📍In the C1 admin console, go to Governance › Forms › Configure tenant-wide request settings, set the maximum bulk request entitlement count to a value between 1 and 100 (default is 10), and save.
    • Adds agent classifiers that evaluate private-data sensitivity, untrusted-content exposure, and exfiltration capability at call time, routing matches to a hook, human approval, or outright block — with an Observe mode that logs what a rule would catch before it can deny anything.
    • Adds bulk and individual reprocessing of findings against current transformation and routing rules, with the option to keep or reset manual changes.
    • Adds a configurable bulk request limit (1–100, default 10) for the maximum number of entitlements a user can request at once, set from Governance > Forms.
    • Adds entitlement binding support within the same app (e.g. Salesforce Admin -> Salesforce Editor) via the entitlement binding picker.
    • SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications now honor OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. silent and fresh authentication requests, and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. passive requests, returning an error instead of silently reusing a session when a request cannot be completed safely.

    +6 more show less

    • Adds a guided step-by-step flow for registering MCP servers, covering server identity, hosting location, URL, access policy, tool discovery defaults, and authentication.
    • Duration limits are now shown up front in access request forms, suppressing the Indefinite option when a max is set and auto-adjusting invalid existing durations.
    • Access request flows now surface account suspension status for the target user or the holder of already-granted access.
    • Revokes on connectors that cannot deprovision a given entitlement now route to manual deprovisioning instead of closing the task as Revoked while leaving the source unchanged.
    • The app details page Edit owners picker adds an All users / Owners toggle so admins can see and manage the full owner list, including suspended or filtered-out users.
    • External Slack accounts (including Slack Connect members) can no longer be linked to a C1 user account by email match; notifications for affected users are skipped with a warning rather than silently bound.
    • Adds agent classifiers that evaluate private-data sensitivity, untrusted-content exposure, and exfiltration capability at call time, routing matches to a hook, human approval, or outright block — with an Observe mode that logs what a rule would catch before it can deny anything.
    • Adds bulk and individual reprocessing of findings against current transformation and routing rules, with the option to keep or reset manual changes.
    • Adds a configurable bulk request limit (1–100, default 10) for the maximum number of entitlements a user can request at once, set from Governance > Forms.

    8 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions that reference the bare workflow context, or pass it to a function, no longer compile — expressions must name the step and field directly. Existing running automations are unaffected until next saved, at which point indirect-reference expressions must be rewritten to a static path before saving.
    • !The Owner column, filter, and detail field on findings are renamed Assignee. API consumers, webhook consumers, and SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. integrations reading the Finding object or audit log must update to the new field and event names.
53. changelog Sep 12, 2026 · issue 072

    ConductorOne: Adds four new read-only API endpoints to preview which session policy applies to a given user (by direct assignment, group membership, or tenant default) and list every user a policy applies to — direct and group-conferred, with the conferring group named.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Enable Observe mode on an Agent Classifier to log risky tool-call patterns without blocking, so you can tune rules safely before enforcement.
    📍In the C1 admin console, go to AI Agents › Classifiers, create or edit a classifier, set the mode to Observe, define rules over private-data sensitivity and exfiltration capability, then save — matches are logged but not blocked.
    Raise the bulk request entitlement cap for high-volume provisioning workflows, such as onboarding, where the default of 10 is too restrictive.
    📍In the C1 admin console, go to Governance › Forms, open tenant-wide request settings, and set the max bulk entitlement request limit to a value between 1 and 100.
    • Adds groups[] field and correlation_uid to system log events, plus four new Identity & Access event classes: Account Change, Entity Management, User Access Management, and Group Management; machine identities now reported as Service account type.
    • Adds Agent Classifiers to evaluate AI agent risk at call time — writing rules over private-data sensitivity, untrusted-content exposure, and exfiltration capability, then routing matches to a hook, human approval, or block; includes Observe mode to log what a rule would catch before enforcement.
    • Admins can now set a custom bulk request entitlement limit (1–100, default 10) from Governance > Forms.
    • Registering a new desktop device now auto-creates a full-access API credential displayed on the success screen.
    • MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server registration now uses a guided step-by-step flow covering server identity, hosting location, URL, access controls, tool discovery behavior, and authentication.

    +6 more show less

    • Adds a Resources tab for MCP servers where admins can review and approve static resources and URI templates discovered by C1; each approved resource gets its own read entitlement governed separately from tool and toolset access.
    • Adds a Connector sync failing finding type that triggers after two consecutive failed full syncs and auto-resolves on the next successful sync; can be disabled per tenant.
    • Entitlement binding picker now includes entitlements from the same app, enabling same-app bindings (e.g. Salesforce Admin -> Salesforce Editor).
    • Access-request flows now display the entitlement's duration limit up front and flag when the target user or holder of already-granted access is suspended.
    • Provisioning steps on an entitlement can now be reordered in place (moved up or down directly) without removing and re-adding them.
    • Role mining gains a shared-access threshold setting, pre-action entitlement list fine-tuning, and optional routing of automatic profile enrollments through an access request task for approval.
    • Adds groups[] field and correlation_uid to system log events, plus four new Identity & Access event classes: Account Change, Entity Management, User Access Management, and Group Management; machine identities now reported as Service account type.
    • Adds Agent Classifiers to evaluate AI agent risk at call time — writing rules over private-data sensitivity, untrusted-content exposure, and exfiltration capability, then routing matches to a hook, human approval, or block; includes Observe mode to log what a rule would catch before enforcement.
    • Adds a Resources tab for MCP servers where admins can review and approve static resources and URI templates discovered by C1; each approved resource gets its own read entitlement governed separately from tool and toolset access.
    • Adds a Connector sync failing finding type that triggers after two consecutive failed full syncs and auto-resolves on the next successful sync; can be disabled per tenant.

    7 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions that reference the bare workflow context, or pass it directly to a function, no longer compile — each expression must name the step and field directly. Automations already running are unaffected until next edited and saved, at which point indirect-reference expressions must be rewritten to a static path.
    • !The Owner column, filter, and detail field on Findings are renamed to Assignee in the UI, API, webhooks, and audit log — consumers of the Finding object via API, webhook, or SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. integrations must update field and event-name references accordingly.
54. changelog Sep 11, 2026 · issue 071

    ConductorOne adds agent classifiers, four session-policy API endpoints, system log event classes, and bulk request limits in this release.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Cap the number of entitlements a user can request at once in bulk to reduce over-provisioning risk.
    📍In the C1 admin console, go to Governance › Forms, open tenant-wide request settings, and set Max bulk request entitlements to a value between 1 and 100 (default is 10).
    • Adds four read-only API endpoints to preview which session policy applies to a given user and why (direct assignment, group membership, or tenant default), and to list every user a policy covers — direct and group-conferred, with the conferring group named.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Admins can now set a max number of entitlements a user can request at once in bulk (1–100, default 10) from Governance > Forms.
    • Adds a Resources tab in MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server management where admins can approve static resources and URI templates discovered on a server; each approved resource gets its own read entitlement governed separately from tool and toolset access.
    • SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications now honor OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. requests for silent or fresh authentication and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. passive requests; unsafe requests return an error instead of silently reusing an existing session.
    • Every admin nav section (except one) now has its own dashboard page breaking apart what previously lived in a single Product digest view.
    • Adds four read-only API endpoints to preview which session policy applies to a given user and why (direct assignment, group membership, or tenant default), and to list every user a policy covers — direct and group-conferred, with the conferring group named.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds a Resources tab in MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server management where admins can approve static resources and URI templates discovered on a server; each approved resource gets its own read entitlement governed separately from tool and toolset access.
    • Every admin nav section (except one) now has its own dashboard page breaking apart what previously lived in a single Product digest view.

    2 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions that reference the bare workflow context, or pass it to a function, no longer compile — expressions must name the step and field directly. Any automation using an indirect reference will need to be rewritten to a static path before it can be saved again.
    • !The Owner column, filter, detail field, and corresponding API and webhook field on findings are renamed to Assignee — consumers of the Finding object via API, webhook, or audit log (e.g. a SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents.) must update field and event-name references.
55. changelog Sep 10, 2026 · issue 070

    ConductorOne adds agent classifiers, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. resource governance, new system log event classes, and four session-policy API endpoints.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Cap the number of entitlements users can bulk-request at once to reduce over-provisioning blast radius.
    📍In the C1 admin console, go to Governance › Forms › Configure tenant-wide request settings and set the bulk request limit to a value between 1 and 100 (default is 10).
    • Adds a bulk request limit setting in Governance > Forms letting admins cap the number of entitlements a user can request at once (1–100, default 10).
    • Adds a dedicated tab for MCP static resources and URI templates discovered on a server, where admins can approve them; each approved resource gets its own read entitlement governed separately from tool and toolset access.
    • Adds per-section dashboard pages that break apart what previously lived in the single Product digest view.
    • Adds Graph explorer to the reorganized admin navigation.
    • Supports OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. silent and fresh authentication requests and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. passive requests in SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications, returning an error instead of silently reusing a session when a request cannot be completed safely.

    +6 more show less

    • Adds a guided one-question-at-a-time flow for connecting MCP servers, covering server identity, hosting, URL, access policy, tool discovery behavior, and authentication.
    • Allows entitlement binding within the same app (e.g. Salesforce Admin → Salesforce Editor) via an updated binding picker.
    • Adds role mining improvements: configurable shared-access threshold, pre-action entitlement list refinement, and optional access request task routing for automatic profile enrollments.
    • Findings can now be reprocessed individually, in bulk, or from a finding's detail view against current transformation and routing rules, with the option to keep or reset manual changes.
    • Same-app entitlement bindings are now possible through an updated picker that shows other entitlements from the same app.
    • Revokes on connectors that cannot deprovision a given entitlement now route to manual deprovisioning instead of closing as Revoked with no source change.
    • Adds a bulk request limit setting in Governance > Forms letting admins cap the number of entitlements a user can request at once (1–100, default 10).
    • Adds a dedicated tab for MCP static resources and URI templates discovered on a server, where admins can approve them; each approved resource gets its own read entitlement governed separately from tool and toolset access.
    • Adds per-section dashboard pages that break apart what previously lived in the single Product digest view.
    • Adds Graph explorer to the reorganized admin navigation.
    • Supports OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. silent and fresh authentication requests and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. passive requests in SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications, returning an error instead of silently reusing a session when a request cannot be completed safely.

    +1 more show less

    • Findings can now be reprocessed individually, in bulk, or from a finding's detail view against current transformation and routing rules, with the option to keep or reset manual changes.

    5 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions consisting of the bare workflow context, or that pass it to a function, no longer compile — expressions must reference the step and field directly. Existing running automations are unaffected until next edited and saved, at which point indirect-reference expressions must be rewritten to a static path.
    • !The Owner column, filter, detail field, and corresponding API/webhook field and audit-log event names on findings are renamed to Assignee. Consumers of the Finding object via API, webhook, or SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. audit log must update to the new field and event names.
56. docs update Sep 10, 2026 · issue 070

    ConductorOne adds a dedicated Non-Human Identity (NHINHIA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls.) dashboard covering app registrations, assumable roles, and managed identities.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Quickly find all unowned non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. across your connected apps to close a common access governance gap.
    📍In the ConductorOne console, go to Identities › Overview, select the NHI tab, then apply the filter Has findings to surface every app registration, assumable role, or managed identity without an assigned owner.
    • Adds an Identities > Overview dashboard that surfaces every human and non-human identitynon-human identityA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. discovered across connected sources in one place.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds a dedicated NHI tab scoped to app registrations (Entra enterprise apps, OAuth apps, GitHub Apps, Databricks service principals), assumable roles (AWS IAMAWS IAMAWS Identity and Access Management is Amazon's service for controlling who and what can call AWS APIs, by attaching policies that allow or deny actions on resources. Cyber tools use it to scope credentials to least privilege. roles, GCP workload identitiesworkload identitiesAn identity assigned to software, such as a service or container, rather than a person. It lets the workload authenticate and receive permissions without sharing a user's credentials.), and managed identities.
    • Surfaces separate tabs for Human users, Service, Secrets, and Agents identity categories alongside the new NHI view.
    • Shows type and subtype detail strings (such as entra.enterprise_application or aws.role.lambda), owner, and open findings per identity on the dashboard.
    • Highlights unowned NHIsNHIsA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. as a first-class finding, with links to finding detail views that include evidence, remediation guidance, and activity trail.

    +1 more show less

    • Supports filtering the identity list by app, owner, type, or Has findings, and search by name.

    6 details in Everything

57. Updated navigation (2026-09-10) Sep 10, 2026 · issue 070

    ConductorOne reorganizes admin nav and adds per-section Overview dashboards across all major areas.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Find per-app governance and connector config in one place after the nav reorganization.
    📍In the C1 admin console, go to Apps › Connectors to manage connector configuration alongside app governance.
    Access the new per-section Overview dashboard to replace the single Analytics page.
    📍In the C1 admin console, select any section (e.g., Security, Governance, Identities) and click its Overview page to view the section-scoped dashboard.
    • Adds a new Platform section housing Policies, Forms, Task log, and Integrations (excluding Connectors).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds a per-section Overview dashboard for every nav section except Apps, replacing the single Analytics page.
    • Moves Connectors into the Apps section to co-locate it with app governance.
    • Moves Automations into the Governance section alongside onboarding and offboarding flows.
    • Renames the Directory nav section to Identities.

    +1 more show less

    • Renames the Identity security nav section to Security, which now also houses Decoys and the Graph explorer.
    • Adds a new Platform section housing Policies, Forms, Task log, and Integrations (excluding Connectors).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
    • Adds a per-section Overview dashboard for every nav section except Apps, replacing the single Analytics page.

    4 more in Everything

58. changelog Sep 9, 2026 · issue 069

    ConductorOne adds agent classifiers, four new session-policy API endpoints, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. resource governance, and four new system-log event classes.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Cap the number of entitlements a user can request at once to reduce over-provisioning risk across the tenant.
    📍In the ConductorOne console, go to Governance › Forms, locate Bulk request limit, and set a value between 1 and 100 (default is 10).
    • Adds four new read-only API endpoints to preview which session policy applies to a given user (by direct assignment, group membership, or tenant default) and list every user a policy covers — including the conferring group.
    • Adds groups[] field and correlation_uid to system log events, plus four new Identity & Access event classes: Account Change, Entity Management, User Access Management, and Group Management.
    • Adds agent classifiers that evaluate private-data sensitivity, untrusted-content exposure, and exfiltration capability at MCP tool-call time, routing matches to a hook, human approval, or block.
    • Adds Observe mode for agent classifiers, which logs what a rule would have caught before it can deny anything.
    • Adds a new tab in MCP server management where admins can approve discovered static resources and URI templates, with each approved resource receiving its own read entitlement.

    +12 more show less

    • Admins can now set a tenant-wide bulk request limit (1–100, default 10) from Governance > Forms.
    • Adds a Connector sync failing finding type that fires after two consecutive failed full syncs and auto-resolves on the next successful sync.
    • Reprocessing findings against current transformation/routing rules is now available individually, in bulk, or from a finding's detail view, with the option to keep or reset manual changes.
    • Entitlement binding picker now shows other entitlements from the same app, enabling same-app bindings (e.g. Salesforce Admin → Salesforce Editor).
    • Registering a new desktop device now auto-creates a full-access API credential shown on the success screen.
    • Adds a guided step-by-step flow for connecting MCP servers, replacing the previous single multi-step form.
    • Improved role mining experience adds a shared-access threshold control, entitlement list fine-tuning before acting, and optional routing of automatic profile enrollments through an access request task.
    • Access-request flows now flag when the target user — or the holder of already granted access — is suspended.
    • Provisioning steps on an entitlement can now be reordered in place (moved up or down) without removing and re-adding them.
    • SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications now honor OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. requests for silent or fresh authentication and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. passive requests, returning an error instead of silently reusing a session when a request cannot be completed safely.
    • The Edit owners picker on an app's details page now includes an All users / Owners toggle so admins can see and manage the complete owner list, including suspended users.
    • Duration limits for entitlements are now displayed up front in request forms; Indefinite is hidden when a max is set and invalid existing durations are auto-adjusted.
    • Adds four new read-only API endpoints to preview which session policy applies to a given user (by direct assignment, group membership, or tenant default) and list every user a policy covers — including the conferring group.
    • Adds groups[] field and correlation_uid to system log events, plus four new Identity & Access event classes: Account Change, Entity Management, User Access Management, and Group Management.
    • Adds agent classifiers that evaluate private-data sensitivity, untrusted-content exposure, and exfiltration capability at MCP tool-call time, routing matches to a hook, human approval, or block.
    • Adds Observe mode for agent classifiers, which logs what a rule would have caught before it can deny anything.
    • Adds a new tab in MCP server management where admins can approve discovered static resources and URI templates, with each approved resource receiving its own read entitlement.

    +3 more show less

    • Admins can now set a tenant-wide bulk request limit (1–100, default 10) from Governance > Forms.
    • Adds a Connector sync failing finding type that fires after two consecutive failed full syncs and auto-resolves on the next successful sync.
    • Reprocessing findings against current transformation/routing rules is now available individually, in bulk, or from a finding's detail view, with the option to keep or reset manual changes.

    9 more in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions that reference the bare workflow context — or pass it to a function — no longer compile; each expression must name the step and field directly. Existing running automations are unaffected until next edited and saved, at which point indirect-reference expressions must be rewritten to a static path.
    • !The Owner column, filter, and detail field on findings are renamed to Assignee. API or webhook consumers of the Finding object and audit-log/SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. pipelines must update to the new field and event names.
59. docs update Sep 4, 2026 · issue 064

    ConductorOne system log gains actor-type classification, API audience field, and four new Identity & Access event classes.

    • Adds actor.user.type_id value 2 / actor.user.type value "admin" to distinguish administrator-tier human actors from non-admin users (type_id 1 / "user") in system log events, effective September 3, 2026.
    • Replaces type_id 99 (Other) with type_id 4 for machine identities (connectors, managed connectors, functions, tunnels) in system log events, effective September 3, 2026.
    • Adds actor.user.groups[] array to system log events listing each actor's platform roles by display name, with a stable uid per role.
    • Adds unmapped.audience field to API Activity events (class_uid 6003) indicating who may call the operation: basic_user, admin_only, super_admin_only, machine_only, or unassigned.
    • Adds audience= to api.request.flags on API Activity events, surfacing the same audience value as a flag.

    +3 more show less

    • API Activity events (class_uid 6003) now include class_name and category_name caption fields.
    • Emits four new Identity & Access event classes: 3001 (Account Change — user account changes and invitations), 3004 (Entity Management — role definition updates), 3005 (User Access Management — role and entitlement grants and revocations), and 3006 (Group Management — group creation and deletion).
    • Privilege-changing operations now emit a paired API Activity event and Identity & Access event linked by a shared metadata.correlation_uid, enabling joins across both event types.

    8 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !Effective September 3, 2026: actor.user.type_id 1 / actor.user.type "user" no longer covers administrator-tier actors — admins now log as type_id 2 / type "admin". Queries filtering type_id = 1 or type = "user" to capture all human activity must be updated to type_id IN (1, 2) or type IN (user, admin).
    • !Effective September 3, 2026: actor.user.type_id 99 (Other) is retired for machine identities; connectors, functions, and tunnels now log as type_id 4. Automation filtering type_id = 99 for machine identities must switch to type_id = 4.
    • !Privilege-changing operations now produce two events (one API Activity, one Identity & Access) instead of one — dashboards or metrics based on event volume may double-count unless scoped by class_uid.
    • !Four new class_uid values (3001, 3004, 3005, 3006) are emitted; pipelines with a fallthrough or default handler that routes by class_uid will misroute these events unless they are handled explicitly.
    • !Most system log events are larger under the updated schema; volume-based ingest metering costs will increase.
60. docs update Sep 4, 2026 · issue 064

    ConductorOne Findings adds a new Connector sync failing detection and renames the Assignee field with updated API identifiers.

    • Adds assignee_identity_user_id as the replacement field on the Finding object for API and webhook consumers, superseding the deprecated computed_owner and assigned_owner fields.
    • Adds ASSIGNEE_CHANGED audit log / SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. event (with old_assignee / new_assignee keys) replacing OWNER_CHANGED (with old_owner / new_owner keys) for finding ownership changes — only newly emitted events use the new name.
    • New Connector sync failing detection fires when a connector records two consecutive failed full syncs, and auto-resolves on the next successful sync.
    • Assignee appears as its own column and filter on the Findings list, and as a detail field on individual findings.

    4 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !computed_owner and assigned_owner fields on the Finding object are deprecated and no longer populated — API and webhook integrations must switch to assignee_identity_user_id.
    • !The audit log / SIEM event previously emitted as OWNER_CHANGED (with old_owner / new_owner keys) is now emitted as ASSIGNEE_CHANGED (with old_assignee / new_assignee keys); historical rows are unaffected, but any downstream SIEM rules or parsers matching OWNER_CHANGED will stop firing for new events.
61. docs update Sep 4, 2026 · issue 064

    ConductorOne adds agent classifiers, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. resource governance, four new system log event classes, and four new session-policy API endpoints.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Cap the number of entitlements a user can request at once in a bulk request to reduce over-provisioning risk.
    📍In the ConductorOne console, go to Governance › Forms › Configure tenant-wide request settings and set the bulk request limit to a value between 1 and 100 (default is 10).
    • New groups[] field and correlation_uid in the system log link paired privilege-change events and expose each actor's platform roles; four new Identity & Access event classes added: Account Change, Entity Management, User Access Management, and Group Management.
    • Four new read-only API endpoints let admins preview which session policy applies to a given user (by direct assignment, group membership, or tenant default) and list every user a policy covers.
    • Admins can now set a maximum bulk-request entitlement limit (1–100, default 10) from Governance > Forms > Configure tenant-wide request settings.
    • Adds agent classifiers that evaluate private-data sensitivity, untrusted-content exposure, and exfiltration capability at call time, routing matches to a hook, human approval, or block; includes an Observe mode that logs what a rule would catch before enforcing it.
    • MCP resource governance: newly discovered static resources and URI templates appear on a dedicated tab for admin approval; each approved resource gets its own read entitlement governed separately from tool and toolset access.

    +4 more show less

    • Guided step-by-step flow for registering MCP servers replaces the single multi-step form, walking through server identity, hosting location, URL, access policy, tool discovery defaults, and authentication one question at a time.
    • Role mining improvements: configurable shared-access threshold, entitlement list fine-tuning before acting, and optional routing of automatic profile enrollments through an access request task for approval.
    • SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications now honor OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. silent/fresh authentication requests and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. passive requests; unsafe requests (e.g. SAML forced authentication without verifiable upstream evidence) return an error instead of silently reusing a session.
    • Revokes on connectors incapable of deprovisioning a given entitlement now route to manual deprovisioning instead of falsely closing as Revoked.

    9 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions consisting of the bare workflow context, or that pass it to a function, no longer compile — expressions must reference the step and field directly. Already-running automations are unaffected until next edited and saved, at which point indirect-reference expressions must be rewritten to a static path.
    • !The Owner column, filter, and detail field on findings are renamed Assignee. API or webhook consumers of the Finding object and audit-log/SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. pipelines must update to the new field and event names.
62. docs update Sep 4, 2026 · issue 064

    ConductorOne now sends expiring-access reminders at 30 days, 14 days, 3 days, 1 day, and 1 hour before a time-limited grant ends.

    ⤢ ⤢ ⤢ ⤢
63. Product digest: September 4, 2026 Sep 4, 2026 · release notes

    Adds MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. resource governance, agent classifiers with Observe mode, reprocessable findings, system log Identity & Access events, and session policy blast-radius APIs.

    └──▷ HOW TO FIND IT
    Use Observe mode to safely test agent classifier rules against live traffic without blocking legitimate calls.
    📍In Agent classifiers, enable Observe mode and review logs for matches on private-data sensitivity or exfiltration capability before switching to block.
    Reprocess existing findings against updated organization rules without creating new duplicates.
    📍In Findings detail view, select Reprocess against current rules and choose to keep or reset manual changes.
    • Agent classifiers evaluate private-data sensitivity, untrusted-content exposure, and exfiltration capability at call time, routing matches to a hook, human approval, or block via Observe mode
    • New Resources tab for reviewing and approving MCP server static resources and URI templates, granting separate read entitlements
    • Reprocess findings against current transformation/routing rules individually, in bulk, or from the detail view with optional reset of manual changes
    • New Connector sync failing finding type triggered by two consecutive failed full connector syncs
    • System log adds four new Identity & Access event classes: Account Change, Entity Management, User Access Management, and Group Management, plus a groups[] field for platform roles and shared_correlation_uid linking privilege change events

    +1 more show less

    • New read-only API endpoints to preview which session policy applies to a user (direct, group, or tenant default) and list all affected users

    6 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions referencing workflow context indirectly no longer compile; must name the step and field directly in new saves.
    • !The Owner column, filter, and detail field on findings are now labeled Assignee; API/hook consumers must update to use the Assignee field/event-name.
64. September 4, 2026 (2026-09-03) Sep 3, 2026 · issue 064

    ConductorOne adds AI agent classifiers with Observe mode, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. resource governance, reprocessable findings, and four new system log event classes.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Set a tenant-wide bulk request limit so users cannot request more than 5 entitlements at once in a single bulk request.
    📍In the ConductorOne console, go to Governance › Forms › Configure tenant-wide request settings, then set Max entitlements per bulk request to 5 (valid range: 1–100, default: 10).
    • Adds correlation_uid field to system log events linking paired events produced by a privilege change, plus a groups[] field listing each actor's platform roles at the time of the event.
    • Adds four new system log Identity & Access event classes: Account Change, Entity Management, User Access Management, and Group Management.
    • Adds four new read-only API endpoints to preview which session policy applies to a given user and why (direct assignment, group membership, or tenant default), and to list every user a policy covers — direct and group-conferred — with the conferring group named.
    • Introduces Agent classifiers for AI access profiles: write rules over private-data sensitivity, untrusted-content exposure, and exfiltration capability, then route matches to a hook, human approval, or a block.
    • Adds Observe mode for agent classifiers, which logs what a rule would have caught before it can deny anything — safe rollout before enforcement.

    +16 more show less

    • New Resources tab in MCP server management where admins can approve static resources and URI templates discovered on a server; each approved resource gets its own read entitlement governed separately from tool and toolset access.
    • Adds guided step-by-step flow for registering an MCP server (server identity, hosting, URL, access policy, tool discovery behavior, authentication) replacing a single multi-step form.
    • Adds bulk request limit configuration under Governance > Forms — admins can set the max number of entitlements a user can request at once in bulk (range 1–100, default 10).
    • Findings can now be reprocessed against current transformation and routing rules — individually, in bulk, or from a finding's detail view — with the option to keep or reset manual changes.
    • New Connector sync failing finding type raised after two consecutive failed full syncs on a connector; auto-resolves on the next successful sync and can be disabled per tenant.
    • Registering a new desktop device now auto-creates a full-access API credential displayed on the success screen.
    • Role mining adds a shared-access threshold control to focus results on entitlements a cohort actually has in common, plus optional routing of automatic profile enrollments through an access request task for approval.
    • Access request forms now display an entitlement's duration limit up front — Indefinite is hidden when a max is set, and invalid existing durations are auto-adjusted.
    • Access request flows now flag when the target user or holder of already granted access is suspended.
    • Entitlement binding picker now shows entitlements from the same app, enabling same-app bindings (e.g. Salesforce Admin -> Salesforce Editor).
    • Provisioning steps on an entitlement can now be reordered in place (move up/down) without removing and re-adding them.
    • Machine identities (connectors, tunnels, functions) are now reported in system log events with the standard Service account type instead of Other.
    • SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications now honor OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. requests for silent or fresh authentication and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. passive requests; a request that cannot be completed safely returns an error instead of silently reusing an existing session.
    • App details page Edit owners picker adds an All users / Owners toggle so admins can see and manage the complete owner list including suspended or filtered-out users.
    • The Access profile enrollments chart on the AI & MCP analytics tab now counts only enrollments in profiles that include at least one MCP tool or toolset.
    • External Slack accounts (including Slack Connect members) can no longer be linked to a C1 user account by email match; affected users have notifications skipped with a warning instead.

    21 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !Automation expressions that consist of the bare workflow context, or that pass it to a function, no longer compile — expressions must reference the step and field directly. Already-running automations are unaffected until next edited and saved, at which point indirect-reference expressions must be rewritten to a static path before saving.
    • !The Owner column, filter, and detail field on the Finding object (API, webhook, and audit log) are renamed to Assignee — consumers feeding findings via API, webhook, or a SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. must update field and event-name references.
65. docs update Aug 30, 2026 · issue 059

    ConductorOne AI Assistant can now explain access history and grant reasons for users and teams.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Investigate unexpected access changes for a team during an incident or audit window.
    📍In the ConductorOne AI Assistant, type: 'What access changed for the Finance team in the last 30 days?'
    Understand why a specific user has an entitlement before making a revocation decision.
    📍In the ConductorOne AI Assistant, type: 'Why does [user] have access to [entitlement]?'
    • Enables querying access history via the AI Assistant — ask which grants and revocations occurred over a time window for a team or user (e.g., 'what access changed for the Finance team in the last 30 days?'), ordered by most recent.
    • Enables asking the AI Assistant why a user holds a specific entitlement, returning the source of the grant — access request, group automation, catalog membership, uplift, or similar.

    2 details in Everything

66. docs update Aug 30, 2026 · issue 059

    ConductorOne adds smart revocation, campaign review columns, access-history AI answers, push-rule execution history, and app-management APIs

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Select a specific C1 tenant when registering an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server, now that --url replaces the broken --local flag on this command.
    copy

    $ c1i mcp servers register --url https://example.conductor.one <other-flags>

    • Adds --url flag (replacing the shadowed --local flag) on mcp servers register, mcp servers test-connection, and mcp servers update-credentials commands so the C1 tenant URL can now be selected via flag rather than requiring C1I_URL or ~/.c1i.yaml.
    • New public API endpoints to list connector-discovered applications and their managed state, and to promote an unmanaged application to managed (returns 409 on repeat promotion and exposes the managed application ID in the response binding).
    • Adds Smart Revocation (off by default, set per app): traces access to upstream sources — direct grants, nested groups, role hierarchies, delegated provisioning — and revokes the chain, not just the surface entitlement; shows a downstream-effects banner before an approver commits.
    • Adds a grant-source filter on revoke steps to scope a revoke to only direct grants.
    • Ask C1 AI assistant can now answer access-history questions: grants and revocations over a time period (ordered by most recent) and the reason behind each grant (access request, group automation, catalog membership, uplift, etc.).

    +1 more show less

    • Every attribute push rule now records execution history — accessible from the Push Rule page (all connectors) or via View execution history in a push rule's row menu (per connector) — logging who was pushed, when, which attributes changed, and failure reasons from the target system.

    6 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !On mcp servers register, mcp servers test-connection, and mcp servers update-credentials, the external MCP server URL flag is now --url instead of --local; scripts or configs passing --local to set the tenant URL on these commands will stop working.
67. changelog Aug 26, 2026 · issue 057

    ConductorOne adds smart revocation chains, Copilot Studio/Gemini MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. integrations, push-rule execution history, and c1i v0.5.0 with new CLI commands.

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Register a new external MCP server using the renamed --server-url flag and the global --url tenant selector.
    copy

    $ c1i --url https://<tenant>.conductor.one mcp servers register --server-url https://my-mcp-server.example.com --name my-governed-tools

    List all tools available through the C1 MCP gateway to verify governed tool exposure before wiring up a Copilot Studio or Gemini agent.
    copy

    $ c1i --url https://<tenant>.conductor.one mcp gateway list-tools

    • Adds --server-url flag on mcp servers register, mcp servers test-connection, and mcp servers update-credentials commands in c1i, replacing the previous --url flag on those subcommands so the global --url can now correctly select the C1 tenant.
    • Adds apps create, apps delete, and apps set-owners subcommands to c1i v0.5.0, rounding out app-management API coverage from the CLI.
    • Adds auth token subcommand to c1i v0.5.0.
    • Adds mcp gateway list-tools and mcp gateway call subcommands to c1i v0.5.0.
    • Adds policies command family to c1i v0.5.0, guarding against a silent deny-all default.

    +16 more show less

    • New public API endpoints to list connector-discovered applications and their managed state, and to promote an unmanaged application to managed (returning HTTP 409 on a repeat promotion with the managed application ID readable from the binding).
    • Each record returned by the public API's past-grants search now includes a unique id field, making individual grant history entries directly addressable.
    • Adds Grant source setting (value: Direct) on automation revoke steps to scope revocation to only direct grants, preventing collateral removal of group-inherited access.
    • New Revoke inherited access per-app setting enables smart revocation that traces access to upstream sources — direct grants, nested groups, role hierarchies, delegated provisioning — before an approver commits; off by default with a downstream-effects banner shown pre-confirmation.
    • Push rules now record full execution history per connector: accessible via the Push rule page (all connectors) or View execution history in a push rule's row menu (single connector), with each entry showing who was pushed, when, which attributes changed, and the target system's rejection reason on failure.
    • Automation circuit-breaker system-log events now include the standard OCSFOCSFAn open schema standard, maintained by the OCSF community under CISA and industry backing, that defines a common event format so cyber tools can share and ingest log data without per-source normalization work. time field, enabling direct SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. ingestion without a parsing workaround.
    • Campaign admins can now configure up to 10 reviewer-visible app user attribute columns (job title, department, cost center, etc.) in the access review table, available once the campaign's Default access review view is set to something other than None.
    • MCP servers are now searchable from global search (Cmd+K) by name, with recently opened servers appearing under Your recent searches.
    • Supports connecting Copilot Studio agents to the C1 MCP gateway, with every tool call authorized and logged under the calling user's identity and publishable to Microsoft 365 Copilot and Teams.
    • Supports Gemini Enterprise as a C1 MCP gateway tool source, with every tool call attributed to the individual user rather than a shared service account.
    • C1AI assistant can now answer access history queries — grants and revocations over a time period ordered by most recent — not just current access state.
    • C1AI assistant now returns the reason behind each grant (access request, group automation, catalog membership, uplift, etc.) alongside who has access.
    • An app's task list tab can now be filtered to show Request tasks, and the Open revocations stat card on the app's Overview tab links directly to that filtered view.
    • Any assigned reviewer can now open a read-only entitlement/resource details drawer from within a review task without requiring additional permissions.
    • Pages containing role-hidden information now display a toast — 'Some information on this page is hidden based on your permissions' — instead of showing a blank field with no indication.
    • c1i v0.5.0 list commands now emit properly-typed Booleans, counts, and nullable fields instead of stringified or inconsistent JSON, and all responses map onto a single predictable exit-code scheme.

    21 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !c1i now requires a full https:// tenant URL; the bare short-name shortcut is retired and a non-HTTPSHTTPSHTTP carried over an encrypted TLS connection. It protects traffic in transit and lets a browser verify the server's certificate. URL is rejected instead of silently upgraded.
    • !On mcp servers register, mcp servers test-connection, and mcp servers update-credentials, the external MCP server URL flag is renamed from --url to --server-url; scripts or configs passing --url to select the MCP server on these commands will break.
68. docs update Aug 22, 2026 · issue 053

    ConductorOne adds a native MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. gateway integration for Google Gemini Enterprise with per-user OAuth attribution and policy enforcement.

    └──▷ TRY IT
    Check and disable the org policy that blocks custom MCP data connectors before creating the Gemini Enterprise data store.
    copy

    $ # Check current effective policy
    gcloud org-policies describe \
      discoveryengine.managed.disableCustomMcpServerConnector \
      --project=YOUR_PROJECT_ID --effective
    
    # Save override to policy.yaml, then apply
    cat > policy.yaml <<'EOF'
    name: projects/YOUR_PROJECT_ID/policies/discoveryengine.managed.disableCustomMcpServerConnector
    spec:
      rules:
        - enforce: false
    EOF
    gcloud org-policies set-policy policy.yaml

    Enable all six required Google Cloud APIs for the Gemini Enterprise MCP integration in one command.
    copy

    $ gcloud services enable \
      discoveryengine.googleapis.com \
      connectors.googleapis.com \
      integrations.googleapis.com \
      secretmanager.googleapis.com \
      iam.googleapis.com \
      orgpolicy.googleapis.com \
      --project=YOUR_PROJECT_ID

    Confirm the C1 OAuth client metadata and discover authorization/token endpoints for your tenant before configuring the Gemini Enterprise data store.
    copy

    $ # Confirm the client metadata document resolves
    curl -s https://YOUR_TENANT.conductor.one/auth/v1/client-metadata/gemini-enterprise
    
    # Discover authorization_endpoint and token_endpoint
    curl -s https://YOUR_TENANT-mcp.conductor.one/.well-known/oauth-authorization-server

    • New ConductorOne MCP gateway integration for Google Gemini Enterprise routes tool calls through https://<your-tenant>-mcp.conductor.one/v1, attributing each call to the individual user's identity rather than a shared service account.
    • Publishes a pre-registered OAuth 2.0OAuth 2.0An open authorization framework (IETF RFC 6749) that lets a service grant limited API access to a third party without sharing credentials, giving cyber tools a standard way to authenticate against external platforms. client metadata document at https://<your-tenant>.conductor.one/auth/v1/client-metadata/gemini-enterprise, eliminating secret rotation — the client_id is the document URL itself and uses PKCEPKCEProof Key for Code Exchange, an OAuth safeguard that binds an authorization code to the client that requested it. An intercepted code cannot be redeemed without the client's matching verifier. with no client secret.
    • Exposes authorization_endpoint (https://<tenant>.conductor.one/auth/v1/authorize) and token_endpoint (https://<tenant>.conductor.one/auth/v1/token) discoverable via https://<tenant>-mcp.conductor.one/.well-known/oauth-authorization-server.
    • Requires the constraints/discoveryengine.managed.disableCustomMcpServerConnector Google Cloud organization policy to be overridden at project level (enforcement set to Off) before a custom MCP data store can be created.
    • Requires six Google Cloud APIs to be enabled: discoveryengine.googleapis.com, connectors.googleapis.com, integrations.googleapis.com, secretmanager.googleapis.com, iam.googleapis.com, and orgpolicy.googleapis.com.

    +5 more show less

    • Supports an optional constraints/discoveryengine.allowedEgressFqdns egress allowlist — add <your-tenant>-mcp.conductor.one and <your-tenant>.conductor.one (hostnames only) if your org restricts outbound connector traffic.
    • Gemini Enterprise MCP data connector requires the Authorization URL Parameters field set to &resource=https://<your-tenant>-mcp.conductor.one/v1; omitting this causes C1 to reject the issued token.
    • OAuth scopes for the connector must be space-separated: openid profile email offline_access.
    • AI client connections are visible in C1 under AI access management > AI clients, and users can review their own connections at profile menu > AI & API > AI connections.
    • Tool calls are gated by C1 toolsets — users with no assigned toolset see no tools; access profiles control which MCP actions each user can invoke.

    10 details in Everything

69. docs update Aug 22, 2026 · issue 053

    ConductorOne adds smart revocation, policy reuse, Copilot Studio/Gemini MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. integrations, push-rule execution history, and c1i v0.5.0 new commands.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Scope an automation's revoke step to direct grants only, so group-inherited access is left intact during automated cleanup.
    📍In the automation editor, open the revoke step, find the Grant source setting, and select Direct.
    List available MCP gateway tools via the c1i CLI to verify which governed tools are callable through the gateway.
    copy

    $ c1i mcp gateway list-tools --server-url https://<tenant>.conductor.one

    • Adds Grant source setting (Direct) on automation revoke steps to scope automated revocations to only direct grants, preventing collateral removal of group-inherited access.
    • Adds apps create, apps delete, apps set-owners, auth token, mcp gateway list-tools, mcp gateway call, and a policies command family to the c1i CLI (v0.5.0).
    • Public API can now list connector-discovered applications, inspect their managed state, and promote an unmanaged application to managed — returning HTTP 409 on a repeat promotion and exposing the managed application ID in the binding response.
    • Each record returned by the public API's past-grants search now includes a unique id field for direct addressability of individual grant history entries.
    • Adds Revoke inherited access toggle on an app's details page — when enabled, a revoke traces and removes upstream sources (direct grants, nested groups, role hierarchies, delegated provisioning) instead of stopping at the first result; shows a downstream-effects banner before an approver commits.

    +11 more show less

    • Push-rule execution history is now available at Directory sources > Executions (all connectors) or via Show execution history in a push rule's row menu — each entry records who was pushed, when, which attributes changed, and the rejection reason on failure.
    • Supports Copilot Studio as a C1 MCP gateway client, with every tool call authorized and logged under the calling user's identity and publishable to Microsoft 365 Copilot and Teams.
    • Supports Gemini Enterprise as a C1 MCP gateway tool source, with every tool call attributed to the individual user rather than a shared service account.
    • MCP servers are now searchable via global search (Cmd+K), with recently opened servers surfaced in Your recent searches.
    • Automation circuit-breaker system-log events now include the standard OCSFOCSFAn open schema standard, maintained by the OCSF community under CISA and industry backing, that defines a common event format so cyber tools can share and ingest log data without per-source normalization work. time field for direct SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. ingestion without a parsing workaround.
    • Policies can now be created local to a single app or entitlement via inline creation alongside the global option on each configuration slot.
    • A policy rule or baseline can now reference another policy of the same type, enabling a shared flow (e.g., a manager-then-owner approval chain) to be defined once and reused — updates to the shared policy propagate automatically to all referencing policies.
    • Access review table columns can now include any reviewer-visible app user attributes (job title, department, cost center) — up to 10 per campaign — so reviewers no longer need to hover each row.
    • Task reviewers can open a read-only details drawer for the entitlement or resource under review directly from the task page, with no special role required.
    • An app's task tab can now be filtered to Request tasks, and the Open revocations stat card on the app's Overview tab links directly to that filtered view.
    • Pages that hide information due to role restrictions now display a toast — 'Some information on this page is hidden based on your permissions' — instead of showing blank fields with no indication.

    16 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !c1i v0.5.0 now requires a full https:// tenant URL — bare short-name shortcuts are rejected, and a non-HTTPSHTTPSHTTP carried over an encrypted TLS connection. It protects traffic in transit and lets a browser verify the server's certificate. URL is rejected instead of silently upgraded.
    • !The mcp gateway command's --url flag has been renamed to --server-url; any scripts or aliases using --url with mcp gateway will break.
70. August 21, 2026 (2026-08-21) Aug 21, 2026 · issue 053

    ConductorOne adds smart revocation chain tracing, automation grant-source scoping, policy reuse, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. gateway integrations, and a major c1i CLI expansion.

    ⤢ ⤢ ⤢ ⤢
    └──▷ TRY IT
    Call a governed MCP tool through the c1i CLI to verify the gateway is reachable and the tool executes under the caller's identity.
    copy

    $ c1i mcp gateway call --server-url https://<tenant>.conductor.one/mcp <tool-name> '{"arg": "value"}'

    Scope an automation revoke step to direct grants only, so group-inherited access is not collaterally removed during cleanup.
    📍In the automation editor, open the revoke step, then set Grant source to Direct instead of All.
    • Adds apps create, apps delete, apps set-owners, auth token, mcp gateway list-tools, mcp gateway call, and a policies command family to the c1i CLI, rounding out API coverage for agent-focused automation.
    • Adds --server-url flag to the c1i mcp gateway command, replacing the former --url flag, so the global --url flag consistently selects the C1 tenant across all commands.
    • New Grant source setting (All or Direct) on automation revoke steps lets automations scope revocations to only direct grants, avoiding collateral removal of group-inherited access.
    • Automation circuit-breaker system-log events now include the standard OCSFOCSFAn open schema standard, maintained by the OCSF community under CISA and industry backing, that defines a common event format so cyber tools can share and ingest log data without per-source normalization work. time field, enabling SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. ingestion without a custom parsing workaround.
    • Public API can now list connector-discovered applications, inspect their managed state, and promote an unmanaged application to managed; returns 409 on a repeat promotion and exposes the binding to recover the managed application ID.

    +14 more show less

    • Each record returned by the public API's past-grants search now includes a unique id field, making individual grant history entries directly addressable.
    • Smart revocation traces access through direct grants, nested groups, role hierarchies, and delegated provisioning — enable Revoke inherited access per-app on the app details page; off by default, with a pre-commit downstream-effects banner.
    • Campaign admins can now display app-user attributes (job title, department, cost center) as dedicated columns in the access review table — up to 10 per campaign.
    • Task reviewers can open a read-only entitlement/resource details drawer directly from the task page, with no special role required.
    • An app's Requests tab can now be filtered to Request or Revoke tasks; the Open revocations stat card on the app Overview tab links directly to that filtered view.
    • Policies can now be created local to a single app or entitlement, with inline creation alongside the global option on each configuration slot.
    • A policy rule or baseline can now reference another policy of the same type, enabling a shared approval flow (e.g. manager-then-owner chain) to be updated once and automatically reflected everywhere it is reused.
    • Copilot Studio agents can now call IT-approved tools through the C1 MCP gateway, with every call authorized and logged under the calling user's own identity, publishable to Microsoft 365 Copilot and Teams.
    • Gemini Enterprise can now call the C1 MCP gateway as a tool source, with every tool call attributed to the individual user rather than a shared service account.
    • MCP servers are now searchable from global search (Cmd+K), with recently opened servers appearing under Your recent searches.
    • Push rule executions now appear in an execution history — accessible via Directory sources > Executions for all connectors or Show execution history in a push rule's row menu — recording who was pushed, when, which attributes changed, and the reason for any failure.
    • When a page contains information hidden by the user's role, C1 now displays a toast: 'Some information on this page is hidden based on your permissions.'
    • Task lists and access review tables now size columns to fit content and scroll horizontally instead of compressing all columns to the same width; long entitlement names truncate with an ellipsis and reveal the full name on hover.
    • c1i list commands now emit properly-typed Booleans, counts, and nullable fields instead of stringified or inconsistent JSON, and all responses map onto a single predictable exit-code scheme.

    19 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !c1i now requires a full https:// tenant URL — bare short-name shortcuts are rejected, and a non-HTTPSHTTPSHTTP carried over an encrypted TLS connection. It protects traffic in transit and lets a browser verify the server's certificate. URL is rejected instead of silently upgraded.
    • !The mcp gateway command's --url flag is renamed to --server-url; any script passing --url to mcp gateway must be updated.
71. Product digest: August 21, 2026 Aug 21, 2026 · release notes

    ConductorOne adds smart revocation for inherited access, C1 AI assistant history queries, c1i v0.5.0 with new commands, and API support for connector app management.

    • Revoke inherited access feature now traces and revokes upstream grant sources (groups, roles) instead of just direct entitlements.
    • Access review campaigns can now configure custom columns for reviewer-visible attributes (up to 10 per campaign).
    • C1 AI assistant Access history query supports asking for grants/revocations over a time period and the reason why access exists.
    • c1i CLI v0.5.0 adds apps create/delete/set-owners, auth token, mcp gateway list-tools/call, and policies commands.
    • Public API exposes endpoints to list connector-discovered apps, promote unmanaged to managed, and search grant history with unique IDs.

    5 details in Everything

    └──▷ BREAKING ON UPGRADE

    • !c1i now requires a full https:// tenant URL; the bare short-name shortcut has been retired.
    • !On mcp servers register, mcp servers test-connection, and mcp servers update-credentials, the external MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server URL flag is now --server-url instead of --url.
72. docs update Aug 18, 2026 · issue 049

    ConductorOne adds Findings: automated detection, triage, transformation rules, and routing rules for identity risk (early access).

    └──▷ HOW TO FIND IT
    Automatically escalate every decoy-credential finding to Critical and notify the security team the moment it is created, so compromise signals never wait in a queue.
    📍1. Go to Identity security › Findings, click the gear icon, toggle on Decoy credential used, and Save changes. 2. Go to Findings › Transformation rules › Create rule. Set Finding type to Decoy credential used. Add a Set severity transform, choose Critical, and save. 3. Go to Findings › Routing rules › Create rule. Set Finding type to Decoy credential used. Set Action to No action. Add a Trigger automation dispatcher pointing to your team notification automation. Set Order lower than any broad catch-all rule, then save.
    Suppress low-severity noise automatically so analysts only see findings that require attention, using a broad catch-all routing rule placed last in the order.
    📍1. Go to Findings › Routing rules › Create rule. 2. Leave App and Finding type unset (Match all). 3. Add a CEL match condition: finding.severity == FINDING_SEVERITY_LOW 4. Set Action to Suppress. 5. Assign an Order value (e.g. 9000) higher than all specific rules so it only fires when nothing else matches. 6. Save the rule.
    • Adds a new Findings module (early access) under Identity security > Findings that continuously monitors for identity risk conditions — unowned service accounts, exposed credentials, misclassified identities, decoy credential use, and more — on every connector sync.
    • Adds per-finding actions — Accept risk (requires justification and expiration date), Suppress (permanent, requires reason), Snooze (1, 7, 30, or 90 days or custom date, requires reason), and Remediate (where supported) — accessible from the finding detail view.
    • Adds bulk actions from All findings: select up to all findings matching the current filter and apply Snooze, Suppress, Accept risk, or Reopen in one step; the confirmation dialog shows how many of the selected findings each action can change.
    • Adds Transformation rules (at Findings > Transformation rules) that reshape a finding's data before routing acts on it — match on finding.severity, finding.state, finding.risk_score, or finding.annotations via CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expression, then set severity or add/remove annotations; all matching rules run in ascending Order.
    • Adds Routing rules (at Findings > Routing rules) that automate triage — match on app, finding type, or a CEL expression, then suppress, snooze, accept, remediate, or dispatch (webhook, function, or automation); only the first matching rule runs, evaluated in ascending Order.

    +1 more show less

    • Finding detail view surfaces an Evidence pane (including confidence score) and an Activity audit trail covering creation, routing-rule evaluation, and evidence updates.

    6 details in Everything

73. docs update Aug 18, 2026 · issue 049

    ConductorOne adds Findings — automated access-risk detection with decoy tripwire credentials and dashboard integration.

    ⤢ ⤢ ⤢ ⤢
    • New Findings capability continuously checks for access risks (unowned service accounts, exposed credentials, misclassified identities) on every connector sync, replacing manual periodic reviews.
    • Supports transformation and routing rules so findings can be triaged, escalated, or resolved automatically without hand-reviewing each result.
    • New Decoys feature lets you plant a tripwire credential that looks real but grants no access — any use attempt raises a Critical finding while the actor sees only a normal authentication failure.
    • Findings surface in the dashboard, with every inventory table now filterable to identities with open findings.

    4 details in Everything

74. Findings and Decoys (2026-08-18) Aug 18, 2026 · issue 049

    ConductorOne adds Findings (automated access-risk detection) and Decoys (tripwire credentials) with triage routing and dashboard integration.

    ⤢ ⤢ ⤢ ⤢
    • New Findings capability automatically checks every connector sync for access risks — unowned service accounts, exposed credentials, misclassified identities — without requiring manual review.
    • New Decoys feature lets teams plant tripwire credentials that look real but grant no access; any use attempt raises a Critical finding while showing the actor only an ordinary authentication failure.
    • Findings surface on the Identities & NHINHIA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. dashboard, with every inventory table now filterable to identities with open findings.

    3 details in Everything

75. docs update Aug 17, 2026 · issue 048

    ConductorOne adds service principals — machine identities with client credentials or secretless OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. workload federation for CI/CD and automation.

    └──▷ TRY IT
    Authenticate a Terraform run to ConductorOne using a pre-existing client ID and Ed25519Ed25519A fast elliptic-curve signature scheme with small keys, the default for new SSH keys and common in code signing. It is not quantum-safe. private key without modifying provider config files.
    copy

    $ export CONDUCTORONE_CLIENT_ID=<your-client-id>
    export CONDUCTORONE_CLIENT_SECRET=<your-ed25519-private-key>
    terraform apply

    Use secretless workload federation in a GitHub ActionsGitHub ActionsGitHub's built-in CI/CD system that runs automated workflows triggered by repository events, letting cyber tools automate testing, building, and releasing without managing separate pipeline infrastructure. workflow — the CI platform supplies the OIDC token so no secret is stored.
    copy

    $ export CONDUCTORONE_CLIENT_ID=<your-federation-trust-client-id>
    export CONDUCTORONE_OIDC_TOKEN=$(cat $ACTIONS_ID_TOKEN_REQUEST_TOKEN)
    cone <subcommand>

    Bypass all token-exchange flows and call the C1 API with a pre-exchanged bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else. — highest-priority auth method.
    copy

    $ export CONDUCTORONE_ACCESS_TOKEN=<your-bearer-token>
    cone <subcommand>

    • Introduces service principals: non-human machine identities assignable C1 roles, usable in scripts, CI/CD pipelines, Terraform runs, and API integrations, separate from any human user account.
    • Supports two authentication methods per service principal: client credentials (client_credentials grant type using a client ID + Ed25519 private key) and secretless workload federation (urn:ietf:params:oauth:grant-type:token-exchange grant type via external OIDC JWTJWTAn open standard (RFC 7519) for encoding claims as a signed, optionally encrypted JSON object passed between parties. Cyber tools use it to verify identity and permissions without querying a central session store.), usable independently or together.
    • Recognizes CONDUCTORONE_CLIENT_ID, CONDUCTORONE_CLIENT_SECRET, CONDUCTORONE_ACCESS_TOKEN, CONDUCTORONE_OIDC_TOKEN, CONDUCTORONE_TENANT_DOMAIN, and CONDUCTORONE_SERVER_URL environment variables across all C1 client tools (Go SDK, Terraform provider, Cone CLI, oidc-token-action), with a defined priority order: CONDUCTORONE_ACCESS_TOKEN first, then CONDUCTORONE_OIDC_TOKEN, then CONDUCTORONE_CLIENT_ID + CONDUCTORONE_CLIENT_SECRET.
    • Workload federation supports GitHub Actions, GitLab CI, HCP Terraform, AWS, and SPIFFESPIFFEAn open standard, maintained by the CNCF, for assigning cryptographic identities to workloads in dynamic infrastructure so services can authenticate each other without passwords or static credentials. as OIDC providers, requiring a publicly accessible JWKSJWKSA JSON Web Key Set, a published collection of cryptographic keys. Applications use it to find the public key needed to verify a token's signature. endpoint; tokens are scoped to individual workflow runs with no secrets to store or rotate.
    • Client credentials enforce a maximum credential lifetime of 180 days with per-credential IP address range restrictions (up to 32 ranges); credential expiration cannot be extended after creation — a new credential must be issued and the old one revoked.

    +1 more show less

    • Each service principal supports multiple credentials or federation trusts simultaneously, has named owners, and appears in the C1 user directory with its own account so entitlements work identically to human users.

    6 details in Everything

76. docs update Aug 17, 2026 · issue 048

    ConductorOne adds admin controls for org-wide notification defaults, per-type locking, and digest scheduling via Settings > Notifications.

    └──▷ HOW TO FIND IT
    Lock task-alert notifications for access reviews so users cannot opt out, ensuring time-sensitive reminders are always delivered.
    📍Navigate to Settings › Notifications. Click Edit in the Notifications area. Locate the relevant notification type under the Requests category. Click the lock icon next to that row. Click Save.
    Set the org-wide C1 digest to deliver a weekly summary via Microsoft Teams every Monday morning.
    📍Navigate to Settings › Notifications. Click Edit in the C1 digest area. Enable the Send C1 digest toggle if not already on. Select Weekly and choose Monday as the delivery day. Select Microsoft Teams as the delivery channel. Click Save.
    • Adds Settings > Notifications admin UI to set organization-wide default notification preferences across Email, Slack, and Teams delivery channels, with bulk checkboxes per channel header.
    • Adds per-notification-type lock icons in Settings > Notifications so admins can prevent users from overriding specific notification settings — locked settings override individual user preferences with the tenant default.
    • Adds C1 digest configuration in Settings > Notifications to set default digest frequency (Daily at 9 AM Pacific on weekdays, or Weekly at 10 AM Pacific on a chosen weekday) and delivery channel (email, Slack, or Microsoft Teams).
    • Slack and Microsoft Teams delivery channel columns appear automatically in the notifications settings table once the respective integrations are configured for the organization.

    4 details in Everything

77. docs update Aug 17, 2026 · issue 048

    ConductorOne launches Cone CLI with entitlement search, request, and revocation commands plus AWS SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. integration

    └──▷ TRY IT
    Request just-in-time privileged AWS access from the command line — Cone creates the approval request and waits for provisioning.
    copy

    $ cone get aws-prod-admin

    Revoke an entitlement you no longer need to reduce your standing access footprint.
    copy

    $ cone drop aws-prod-admin

    Switch between multiple C1 tenant profiles in scripts by setting the active profile via environment variable.
    copy

    $ CONE_PROFILE=prod-tenant cone search

    • New cone search command lists all entitlements currently available to you across applicable access profiles.
    • New cone get <entitlement_alias> command requests access to a named entitlement, automatically creates an approval request if access is not yet granted, and may prompt for justification or access duration based on entitlement settings.
    • New cone drop <entitlement_alias> command creates a revocation request for a held entitlement and deprovisions access after any required review.
    • New cone login <tenant-name or tenant-url> command authorizes the CLI via browser and writes credentials to $HOME/<user name>/.conductorone/config.yaml.
    • Supports multiple named profiles in $HOME/<user name>/.conductorone/config.yaml; active profile selected via $CONE_PROFILE environment variable or --profile flag.

    +3 more show less

    • Integrates with AWS IAMAWS IAMAWS Identity and Access Management is Amazon's service for controlling who and what can call AWS APIs, by attaching policies that allow or deny actions on resources. Cyber tools use it to scope credentials to least privilege. Identity Center to request and retrieve AWS credentials directly from the AWS CLI via Cone.
    • Introduces c1i, a separate CLI designed for machine-readable output and broad API access, aimed at automation, scripts, and AI agent workflows.
    • Available on macOS, Windows, and Linux for x86 and ARMARMA family of RISC processor architectures designed by Arm Holdings and licensed to chip makers worldwide; tools target it to run natively on mobile, embedded, and increasingly server and desktop hardware. platforms; installable via brew install conductorone/cone/cone or direct GitHub release download.

    8 details in Everything

78. docs update Aug 17, 2026 · issue 048

    ConductorOne adds a Microsoft Teams app for approving access requests, tracking open requests, and receiving connector sync alerts without leaving Teams.

    └──▷ HOW TO FIND IT
    Connect your organization's Microsoft Teams directory to ConductorOne so users receive access-request notifications and approvals inside Teams.
    📍In C1, go to Admin › Settings › Notifications. In the Microsoft Teams section, click Connect. Follow the Microsoft prompts and grant Organization.Read.All and User.Read.All permissions. Repeat for each additional Teams directory by signing in as an admin on that directory.
    • New Microsoft Teams app lets users track open access requests, approve or deny requests, and request new access without leaving the Teams interface.
    • Connector sync alerts delivered via Teams: notifies users when a connector has a sync issue or a significant drop in synced resources.
    • Connect the C1 Teams app via Admin > Settings > Notifications > Microsoft Teams section; supports connecting multiple Teams directories by repeating the auth flow per directory.
    • App setup requires Organization.Read.All and User.Read.All Microsoft permissions and the Super Administrator role in C1.
    • Disconnect flow removes the integration from C1 via Settings > Notifications and requires deleting the Enterprise Application from Entra IDEntra IDMicrosoft's cloud-based identity and access management service, formerly Azure Active Directory, used by cyber tools to authenticate users and enforce access policies in Microsoft 365 and Azure environments. to fully revoke permissions.

    5 details in Everything

79. docs update Aug 17, 2026 · issue 048

    ConductorOne launches a Slack app with an AI assistant for access requests, reviews, and org data queries.

    • Adds /c1 request slash command in Slack to submit access requests from anywhere in the workspace.
    • Enables the C1 AI assistant (available via the Chat tab in the C1 Slack app, @C1 mentions in channels, or direct messages) to answer natural-language questions about org access data — who has access to what, entitlement catalog contents, IdPIdPA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems.-sourced org structure, and more.
    • C1 AI assistant can generate downloadable files in CSV, JSON, PDF, or plain text format on request (e.g. 'give me a CSV of everyone with admin access'), with a download link appearing directly in the Slack thread.
    • C1 AI assistant can take action on access data — approve, deny, reassign, or escalate access requests; manage access and entitlements; configure campaigns and conflict monitors; build automations and policies — always requiring explicit user approval before making changes.
    • Adds in-Slack notifications and inline approve/deny actions for access request reviews, eliminating the need to leave Slack.

    +4 more show less

    • Adds in-Slack notifications for manual provisioning and deprovisioning assignments, with the ability to mark them complete without leaving Slack.
    • Adds in-Slack notifications for access review tasks assigned during access review campaigns.
    • Adds a History tab in the C1 Slack app to review past AI assistant conversations, and a New Chat button to start a fresh session.
    • Adds campaign Slack channel integration: campaign owners can create or link a Slack channel from the campaign's Configuration tab > Communications area > Add to Slack channel, automatically adding all campaign owners and assigned reviewers to the channel.

    9 details in Everything

80. docs update Aug 17, 2026 · issue 048

    ConductorOne introduces Automations: custom workflows with triggers, steps, CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expressions, versioning, and an AI assistant.

    ⤢ ⤢ ⤢ ⤢
    • New Automations engine lets Super Admins build custom multi-step workflows triggered by employee lifecycle events (onboarding, offboarding, role transfers, access reviews) via Admin > Automations.
    • Automation triggers and steps are fully configurable: 11 trigger types and 15 step types available, with step reordering and deletion from the setup UI.
    • Supports CEL expressions via the Advanced tab of each step's setup drawer — including the ctx object, {{ }} template syntax, and step-output references using each step's Step ID — to conditionally skip steps or define dynamic values.
    • Automations are versioned (starting at v1), with full version history accessible via the … menu > Show version history and the ability to restore any prior version; publishing a new version does not interrupt in-progress executions.
    • New Executions tab on the Automations page provides a unified cross-organization log of all automation runs, filterable by automation, status, and app, with per-run step-by-step detail views.

    +6 more show less

    • Circuit breaker setting lets admins set a rate limit on any automation — when exceeded, C1 pauses the automation and queues new trigger events for review rather than running them.
    • AI-powered Automation assistant (early access) generates complete automations — trigger, steps, and actions — from a plain-language prompt; accessible via Create with Automation assistant on the Automations page.
    • App-scoped automations allow App Admins to create and manage automations tied to a single application from that app's Automations tab under Admin > Applications.
    • Unused access automations fire when a user has not logged into an app account for a configurable period; templates include Send a notification after 30 days and Revoke access after 45 days, manageable from the Unused access section of each app's Overview page.
    • Requestable automations let users trigger workflows from the self-service catalog with approval flows and form inputs.
    • Functions support extends automations with custom TypeScript logic for use cases beyond built-in steps.

    11 details in Everything

81. docs update Aug 17, 2026 · issue 048

    ConductorOne adds OCSFOCSFAn open schema standard, maintained by the OCSF community under CISA and industry backing, that defines a common event format so cyber tools can share and ingest log data without per-source normalization work.-formatted system logs exportable to S3/Azure Blob with SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents. integration for Splunk, Panther, and CrowdStrike

    └──▷ TRY IT
    Retrieve the full list of C1 API events in YAML (Sigma Detection Format) to build detection rules or audit coverage in your SIEM.
    copy

    $ curl -s 'https://<YOUR C1 TENANT URL>/api/v1/ocsf-events.yaml'

    Set up a system log exporter to continuously push OCSF-formatted logs to an S3 bucket for ingestion into Splunk or CrowdStrike Falcon Next-Gen SIEM.
    📍In C1, go to Settings › System Log › Add exporter. Name the exporter (e.g. System log to S3), select your pre-created S3 datasource (ensure the bucket policy includes s3:PutObject), optionally set a file prefix, choose your output format and compression algorithm, then click Save.
    • New system log feature stores all C1 API activity (end-user and administrative) in OCSF (Open Cybersecurity Schema Framework) format, accessible via API or exported to an external data source.
    • Adds GET <YOUR C1 TENANT URL>/api/v1/ocsf-events.yaml and GET <YOUR C1 TENANT URL>/api/v1/ocsf-events.json endpoints to download the authoritative list of all API events in Sigma Detection Format.
    • New system log exporter, configured under Settings > System Log > Add exporter, syncs logs to an external data source (S3 bucket or Azure Blob container) every four to five minutes, generating files of up to 2.5 MB of uncompressed data per run (~280 files per day).
    • OCSF log events include an activity_id field with values 1 (Create), 2 (Read), 3 (Update), and 4 (Delete) for filtering by API call type.
    • Supports SIEM ingestion into Panther, Splunk, and CrowdStrike Falcon Next-Gen SIEM via S3 bucket connectors; CrowdStrike ingestion uses an OCSF parser to normalize events to CrowdStrike Parsing Standard (CPS).

    5 details in Everything

82. docs update Aug 17, 2026 · issue 048

    ConductorOne adds custom key/value annotations to policies, apps, entitlements, and more — editable via UI, Terraform, and API.

    └──▷ USE IT
    Tag a ConductorOne policy with cost-center and compliance metadata so it can be queried in dashboards and routed in CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expressions.
    hcl copy

    resource "conductorone_policy" "approval_default" {
      display_name = "Default approval"
      annotations = {
        "myteam/cost-center" = "iam-platform"
        "myteam/audit"       = "soc2"
      }
    }

    Update a policy's annotations via the API — setting one key and deleting another — without affecting other existing annotation keys.
    copy

    $ curl -X POST "$C1_URL/api/v1/policies/$POLICY_ID" \
      -H "Authorization: Bearer $TOKEN" \
      -H "Content-Type: application/json" \
      -d '{
        "policy": {
          "id": "'"$POLICY_ID"'",
          "annotations": {
            "myteam/cost-center": "iam-platform",
            "myteam/audit": ""
          }
        },
        "updateMask": "annotations"
      }'

    Route approvals to the right team in a policy approver CEL expression based on an annotation, without maintaining a separate mapping.
    text copy

    entitlement.annotations["team"] == "infrastructure"

    • Adds annotations key/value metadata to ConductorOne objects (policy, app, app entitlement, app resource, access profile, automation, access review template), stored on the object, returned on every read, and tracked in change history.
    • Adds annotations block to the Terraform provider for all supported resources (conductorone_policy, conductorone_app, conductorone_app_entitlement, conductorone_custom_app_entitlement, conductorone_app_resource, conductorone_access_profile, conductorone_automation, conductorone_access_review_template).
    • Terraform provider automatically sets IaCIaCInfrastructure as Code: the practice of defining and provisioning compute infrastructure through machine-readable config files rather than manual processes, giving cyber tools repeatable, version-controlled environment setup. well-known annotation keys managed_by, iac_workspace (from TF_WORKSPACE), and iac_resource_address; iac_tool_version must be set manually.
    • Supports managed_by annotation values terraform, opentofu, pulumi, crossplane, and ansible; objects with IaC keys display a Managed by [tool] chip on their detail page.
    • Adds annotations field to the POST /api/v1/policies/$POLICY_ID (and equivalent per-object) API endpoints with merge semantics — include 'annotations' in updateMask to set/update keys; set a key's value to empty string to delete it.

    +4 more show less

    • CEL expressions in policies, approver rules, and entitlement routing rules can now read entitlement.annotations, role.annotations, and scope.annotations using bracket notation (e.g., entitlement.annotations["team"]) to fork logic on annotation metadata.
    • Adds Take ownership behavior: editing any non-annotation field on an IaC-managed object via the C1 UI clears the IaC keys (managed_by, iac_workspace, iac_resource_address) in the same operation.
    • Adds annotation change history to each object's detail page via the Change history icon, with a distinct badge for changes to IaC management keys.
    • Annotation limits: max 16 keys per object, key length 1–128 characters, value length 0–256 characters, total size ≤ 4 KB per object; keys starting with c1/ are rejected.

    9 details in Everything

83. docs update Aug 17, 2026 · issue 048

    ConductorOne adds External Insights: CrowdStrike and Wiz risk scores surface in access reviews and drive CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. policy automation

    └──▷ USE IT
    Automatically block access requests when CrowdStrike reports a risk score above 70 — useful for enforcing a risk-based access gate in a CEL policy condition.
    text copy

    c1.app_user.v1.GetRiskScore(account, "CrowdStrike").normalized_score > 70

    Deny access when Wiz has flagged CRITICAL or HIGH severity findings, preventing privileged access grants to compromised identities.
    text copy

    c1.app_user.v1.HasSecurityInsightWithSeverity(account, "Wiz Insights", "CRITICAL") || c1.app_user.v1.HasSecurityInsightWithSeverity(account, "Wiz Insights", "HIGH")

    Require a clean posture from both sources before granting sensitive access — CrowdStrike score at or below 30 and no HIGH or CRITICAL Wiz findings.
    text copy

    ("CrowdStrike" in account.risk_score && account.risk_score["CrowdStrike"] <= 30) && !c1.app_user.v1.HasSecurityInsightWithSeverity(account, "Wiz Insights", "HIGH") && !c1.app_user.v1.HasSecurityInsightWithSeverity(account, "Wiz Insights", "CRITICAL")

    • New External Insights feature syncs identity risk data from CrowdStrike Falcon Identity Protection and Wiz Insights into C1, matching scores to identities by email address and surfacing them across access reviews, approvals, and task logs.
    • Enables or disables identity risk score syncing via the Identity Risk Score toggle under Resource capabilities on a connector's Details tab — no additional setup required when a supported connector is already syncing.
    • Exposes CrowdStrike risk scores via account.risk_score (map keyed by source display name, returns integer) and account.risk_scores (list of score objects) in CEL policy conditions for automated access decisions.
    • Exposes Wiz security findings via account.security_insights in CEL, with each insight carrying source, value, and severity fields ('CRITICAL', 'HIGH', 'MEDIUM', 'LOW', 'INFORMATIONAL').
    • Adds CEL helper function c1.app_user.v1.HasSecurityInsight(account) — returns bool true if the account has any issue insights from any source.

    +8 more show less

    • Adds CEL helper function c1.app_user.v1.GetSecurityInsights(account, source) — returns a list of all Wiz issue insights from the named source.
    • Adds CEL helper function c1.app_user.v1.HasSecurityInsightWithSeverity(account, source, severity) — returns bool true if the account has at least one finding at the given severity (case-insensitive comparison).
    • Adds CEL helper function c1.app_user.v1.HasRiskScore(account) — returns bool true if the account has risk scores from any source.
    • Adds CEL helper function c1.app_user.v1.GetRiskScore(account, source) — returns a c1.risk_score.v1 object with a normalized_score integer (0–100; defaults to 0 if missing).
    • Surfaces CrowdStrike password risk factors EXPOSED_PASSWORD and WEAK_PASSWORD in the UI alongside numeric risk scores when password risk ingestion is enabled on the connector.
    • Adds an Insights tab on access review tasks so reviewers see an identity's risk score and risk factors (e.g., STALE_ACCOUNT, WEAK_PASSWORD_POLICY) inline without leaving C1.
    • Adds an Insights column to the task log with a hover summary of risk score and risk factors, plus a link to full details.
    • Surfaces identity risk scores and risk factors to approvers during access request approval tasks before they submit a decision.

    13 details in Everything

84. docs update Aug 17, 2026 · issue 048

    ConductorOne adds an Inventory page with unified Identities, Resources, and Secrets tabs across 8 integrations.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Enable secrets syncing for an existing Okta v2, GitHub v2, AWS v2, Snowflake v2, or Datadog v2 connector to start seeing credentials on the Secrets tab.
    📍In ConductorOne, go to Integrations › Connectors, locate your connector, open its Settings area, click Edit, check Sync secrets, then click Save. To pull data immediately, navigate to the connector's page and click Sync now.
    Track unused or expiring credentials across your environment to identify stale API keys and tokens.
    📍In ConductorOne, go to Identity security › Inventory, click the Secrets tab, locate the credential of interest, and click the associated name in the Identity column to open the identity details page and review the Secrets tab and any active alerts.
    • New Inventory page (under Identity security > Inventory) provides a single-pane-of-glass view of all identities, resources, and secrets synced to ConductorOne.
    • New Identities tab on the Inventory page shows all user, service, and system accounts synced to C1.
    • New Resources tab on the Inventory page lists all resources with associated entitlements and per-resource account-grant counts.
    • New Secrets tab on the Inventory page surfaces sensitive credentials from configured connectors — AWS (access keys), Datadog (API keys), GitHub (API keys), Google Cloud Platform with Google Workspace (API keys, service account keys), Okta (API tokens), Oracle Cloud Infrastructure (API keys), Rootly (secrets), and Snowflake (RSARSAThe oldest widely used public-key algorithm, for both encryption and signatures, with security resting on factoring large numbers. A large quantum computer would break it. public keys).
    • New Sync secrets checkbox in the Settings area of the Okta v2, GitHub v2, AWS v2, Snowflake v2, and Datadog v2 connector setup pages enables secret syncing for those integrations.

    +1 more show less

    • The Secrets tab tracks expiration and last-used date for each credential, and surfaces active alerts for unused tokens or keys on the associated identity's details page.

    6 details in Everything

85. docs update Aug 17, 2026 · issue 048

    ConductorOne adds Access Explorer queries and a Security dashboard for auditing high-risk and orphaned access.

    • New Access Explorer query library at Explore > Search surfaces pre-built queries covering orphaned accounts, inactive accounts, high-risk accounts, standing privileges, past grants, apps/entitlements/resources with deactivated or single owners, and more — each with available filters (by app, user, entitlement, resource type, or risk level).
    • New Security dashboard at Explore > Analytics > Security tab aggregates six high-risk access categories — Orphaned accounts, High-risk accounts, Inactive accounts (past 30 days), High-risk role grants (permanent), High-risk role grants (temporary), and Standing privileges — with drill-through to full query results and report download.

    2 details in Everything

86. docs update Aug 17, 2026 · issue 048

    ConductorOne AI assistant now available to all users, with fast trend charts built from pre-aggregated metrics.

    ⤢ ⤢ ⤢ ⤢
    • Extends the C1 AI assistant to all C1 users, supporting questions and actions in the same language the user writes in.
    • Adds trend charts (grants vs. revocations, review decisions, automation runs over time) built from pre-aggregated metrics, returning exact, traceable numbers without on-the-fly computation.

    2 details in Everything

87. docs update Aug 17, 2026 · issue 048

    ConductorOne MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server list now shows an inline restart button for paused servers.

    ⤢ ⤢ ⤢ ⤢
88. changelog Aug 10, 2026 · issue 041
    ⤢ ⤢ ⤢ ⤢
    • New Identities & NHI dashboard (under the main navigation) provides a unified, filterable, searchable, and exportable inventory of human users, service accounts, API secrets, agents, and other non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. across all connected apps, with credential-to-identity linking and unowned NHINHIA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. flagging.
    • TypeScript typing and autocomplete for @c1/functions-sdk and @c1/test now available in the function editor, surfacing misspelled methods and wrong argument types at edit time rather than runtime.
    • Task details drawer now shows what the reviewed resource has access to inline, so reviewers no longer need to leave the task to gather context.
    • Notification settings updated: 'your access request was created' confirmation is now toggleable per channel (email, Slack, Teams), on by default; provisioning notifications moved under a dedicated section; always-on system notifications no longer shown as a toggle.
89. docs update Aug 8, 2026 · issue 039

    ConductorOne adds a Campaign Assistant that configures access review scope, policy, and owners from plain-language descriptions.

    • New Create with Campaign assistant option lets practitioners describe an access review in plain language (e.g., "Review all Salesforce access granted in the last 90 days") and automatically configures the campaign's policy, owners, and scope.
    • Supports targeting all entitlements of a given type within a specific app via Entitlements > entitlement type selector (e.g., all groups within Google Workspace), in addition to individual entitlement or Application access selections.
    • Adds Custom selectors as a scoping option when building campaigns, enabling more granular entitlement targeting alongside the standard entitlement-type and application-access approaches.

    3 details in Everything

90. docs update Aug 8, 2026 · issue 039

    ConductorOne adds Copilot Studio and ChatGPT/Cursor as supported AI tool integrations.

    • Supports connecting Copilot Studio to ConductorOne, including publishing the agent to Microsoft 365 Copilot.

    1 detail in Everything

91. docs update Aug 8, 2026 · issue 039

    ConductorOne adds NHINHIA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. dashboard, public API for campaign reports, and TypeScript-aware function editor in its August 2026 release.

    ⤢ ⤢ ⤢ ⤢
    └──▷ HOW TO FIND IT
    Review which of your NHIsNHIsA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. are unowned or flagged as risks across all connected apps in one place.
    📍In the ConductorOne console, go to Identities & NHI. Use the Ownership status filter to surface unowned service accounts, secrets, or agents. Export the filtered table for remediation tracking.
    • New Identities & NHI dashboard provides a unified, filterable, searchable, exportable inventory of human users, service accounts, API secrets, agents, and other non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. across all connected apps — with unowned or risky NHIs flagged automatically.
    • Campaign report generation (JSON/CSV/XLSX) is now available via the public API, including for in-progress campaigns — eliminating the manual UI-export-and-transfer step.
    • New Create with Campaign assistant on the Campaigns page lets users describe an access review in plain language and have policy, owners, and scope configured automatically.
    • Request and automation forms now support date fields (with optional earliest/latest limits, fixed or relative to today) and C1-user fields (single- or multi-select, optionally scoped to a specific set of people).
    • The function editor now provides full TypeScript typing and autocomplete for @c1/functions-sdk and @c1/test, surfacing type errors and misspelled methods at edit time.

    +10 more show less

    • C1AI can now generate trend charts (grants vs. revocations, review decisions, automation runs) directly from C1's aggregated metrics.
    • The task details drawer now surfaces what the reviewed resource has access to, inline alongside the review decision.
    • Bulk-revoke is now available for multiple denied reviews from a campaign's denied-reviews tab in a single action.
    • Entitlements can now be removed directly from a campaign or template's scope summary without re-entering scope configuration.
    • Devices list now supports filtering by owner, device type, and full-lifecycle status (Pending, Active, Revoked, Disabled).
    • MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers list now shows an inline Resume button for paused servers.
    • Guided onboarding flow for new tenants covers access profiles, access reviews, and automated access management in a single setup sequence.
    • The Grants tab on an application account now supports configurable columns including resource type, description, grant count, and grant origins (when grant reasons are enabled).
    • Data grid columns are now resizable by dragging, with widths persisted per table and a double-click to reset to default.
    • Provisioning notifications have moved under a dedicated location; the 'your access request was created' confirmation notification can now be toggled per channel (email, Slack, Teams), and always-on system notifications are no longer shown as a toggle.

    15 details in Everything

92. August 7, 2026 (2026-08-08) Aug 8, 2026 · issue 040

    ConductorOne adds an NHINHIA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. dashboard, AI campaign assistant, public API campaign reports, and date/user form fields in its August 7 release.

    ⤢ ⤢ ⤢ ⤢
    • New Identities & NHI dashboard (under Identity security) provides a unified, filterable, searchable, exportable inventory of human users, service accounts, API secrets, agents, and other non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. across all connected apps — with credentials automatically linked to their owning identity and unlinked credentials flagged separately.
    • AI agents (both ConductorOne-managed and app-discovered) now appear as a first-class category in the NHI dashboard, with context on where they live.
    • Bulk-revoke is now available for denied reviews: select multiple denied items from a campaign's Access reviews tab and revoke all at once.
    • Paused MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers now show an inline Resume button in the MCP servers list, replacing the overflow-menu sync action.

    4 details in Everything

93. Product digest: August 7, 2026 Aug 7, 2026 · release notes

    ConductorOne adds Identities & NHINHIA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. dashboard, Campaign Assistant, API-exportable reports, and UI/SDK usability improvements

    └──▷ HOW TO FIND IT
    Inspect an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server list and restart a paused agent directly from the UI.
    📍In the Cyber tooling console, go to MCP servers and click the Resume button next to the paused server.
    • Adds the Identities & NHI dashboard in the Identity security section for unified inventory of human and non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls.
    • Adds Campaign assistant (Create with Campaign assistant) to configure policy, owners, and scope via natural language on the Campaigns page
    • Adds inline Resume button for paused MCP servers in the MCP servers list
    • Enables C1AI trend chart generation for grants, revocations, review decisions, and automation runs
    • Exposes campaign report generation (JSON/CSV/XLSX) via the public API

    +7 more show less

    • Adds guided setup flow for access profiles, access reviews, and automated access management in new tenants
    • Shows resource access paths directly in the access review task details drawer
    • Enables bulk revocation of denied reviews from the campaign’s Access reviews tab
    • Allows removing entitlements directly from the scope summary without opening the scope configuration
    • Adds date fields with earliest/latest limits and C1-user single/multi-select fields to request/automation forms
    • Provides TypeScript typing and autocomplete for @c1/functions-sdk and @c1/test in the function editor
    • Makes data grid columns resizable by dragging, with width persistence per table and double-click reset

    12 details in Everything

94. docs update Aug 4, 2026 · issue 035

    ConductorOne AI assistant now delivers a personalized onboarding plan with agent-assisted setup steps at tenant creation.

    ⤢ ⤢ ⤢
    • C1AI opens automatically on new tenant creation and runs a guided onboarding conversation, asking about industry, organization size, and goals to build a categorized onboarding plan.
    • Detects the identity provideridentity providerA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems. used during SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. sign-up and incorporates it into onboarding step suggestions.
    • Supports agent-assisted onboarding steps — C1AI can execute tasks such as connecting an app directly from the chat panel, without requiring manual navigation.

    3 details in Everything

95. changelog Jul 31, 2026 · issue 031

    ConductorOne adds AI agent policy steps, group/role access reviews, per-user MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. auth, and scope-limited API keys in its July 2026 release.

    └──▷ HOW TO FIND IT
    Restrict a new personal API key to read-only access so it can be shared with a reporting script without granting full account permissions.
    📍In the console, go to your personal API key creation screen and use the Limit Scopes selector to select a read-only role before saving the key.
    Require each user to authenticate with their own credential when connecting to an external MCP server instead of sharing an admin credential.
    📍In the console, open the MCP server's Configure server settings panel, enable per-user authentication, and select the credential type (OAuth2, bearer token, custom header, or basic auth).
    • Assigns AI agent (C1 AI) to workflow steps in request or review policies, with custom instructions, audit-log attribution to a C1 System identity, and reviewable evaluation conversations.
    • Extends request catalog search to surface an Entitlements matching section — showing entitlements that indirectly grant access to the searched resource — plus a View access graph link showing the full grant chain.
    • Adds a Grant types selector to access review campaigns so reviews can now target grants held by groups or roles, not just users (options: Users only, Resources only, Users and resources).
    • Enables custom recurrence intervals (e.g., every two days, every three months) and configurable campaign durations on campaign templates.
    • Renders IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy documents as a collapsible JSON tree via a View policy button on resource-level access reviews and task details, replacing raw unformatted JSON.

    +10 more show less

    • Makes C1 groups requestable — they can be added to a request catalog or access profile and go through the same approval flow as other access.
    • Queues requests for entitlements created ahead of connector sync (e.g., via Terraform), shows the wait reason, and auto-resumes after sync, with a 24-hour fallback to manual provisioning.
    • Adds a Retry failed provisioning action on failed access tasks, preserving previously collected approvals without restarting the task.
    • Enables step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) approvals that simultaneously require a comment, prompting identity-provider verification before completing.
    • Adds View lineage to AI-generated reports, exposing the exact queries behind each number, verification against the execution log, and a copy action for audit workpapers.
    • Renders multiple AI-requested charts as an aligned grid with per-chart export and sources; supports composed reports combining stat cards, trend charts, and data tables in a single response.
    • Adds a Generate action in the CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. editor that converts a plain-English description directly into a CEL expression.
    • Supports per-user authentication on external MCP servers (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth), replacing the shared admin credential model.
    • Adds per-MCP-serverMCP-serverA process that exposes tools, resources, and prompts over Anthropic's Model Context Protocol, letting an AI agent call external APIs or local services through a standardised interface. tool enablement settings, letting admins configure whether tools require manual enablement or are auto-enabled independently of the tenant-wide AI governance default.
    • Allows any user to restrict a personal API key to a narrower role (e.g., read-only) via the Limit Scopes selector at creation time, without requiring Super Admin access.

    15 details in Everything

96. docs update Jul 28, 2026 · issue 028

    ConductorOne MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server registration gains discovery-issue acknowledgment flow, custom tool prefix override, and per-server tool enablement control.

    └──▷ HOW TO FIND IT
    Configure a newly registered MCP server to auto-enable its tools regardless of the tenant default approval policy.
    📍In the ConductorOne console, go to Admin › MCP Servers, select your server, and set Tool enablement to Auto-enable.
    • Adds a pre-registration discovery check that surfaces relaxable authorization server issues (e.g., issuer mismatch, cross-host auth server) for admin acknowledgment before proceeding, and blocks registration entirely for issues that cannot be safely relaxed.
    • Supports an optional tool prefix override on MCP server registration; falls back to the server's declared prefix, then a C1-generated default.
    • Adds per-server tool enablement setting (Require enablement or Auto-enable) that overrides the tenant-wide default tool approval setting for a specific MCP server.

    3 details in Everything

97. docs update Jul 28, 2026 · docs issue 004

    ConductorOne MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server registration gains discovery-issue acknowledgment flow, custom tool prefix override, and per-server tool enablement control.

    └──▷ HOW TO FIND IT
    Configure a newly registered MCP server to auto-enable its tools regardless of the tenant default approval policy.
    📍In the ConductorOne console, go to Admin › MCP Servers, select your server, and set Tool enablement to Auto-enable.
    • Adds a pre-registration discovery check that surfaces relaxable authorization server issues (e.g., issuer mismatch, cross-host auth server) for admin acknowledgment before proceeding, and blocks registration entirely for issues that cannot be safely relaxed.
    • Supports an optional tool prefix override on MCP server registration; falls back to the server's declared prefix, then a C1-generated default.
    • Adds per-server tool enablement setting (Require enablement or Auto-enable) that overrides the tenant-wide default tool approval setting for a specific MCP server.

    3 details in Everything

98. docs update Jul 28, 2026 · issue 028

    ConductorOne AI Assistant gains visual report generation with charts, stat cards, tables, and full data lineage verification.

    └──▷ TRY IT
    Render a side-by-side dashboard of grants trend and users by status directly in the AI Assistant.
    copy

    $ dashboard: grants trend and users by status side by side

    Audit a report's data provenanceprovenanceEvidence recording where a software artifact came from and how it was built, including its source and build process. It helps a consumer check that the artifact matches a trusted build. by opening the lineage drawer and exporting for workpapers.
    📍In the AI Assistant report, click View lineage to open the Data lineage drawer, review Verified/Unverified status and tool calls, then click Copy as Markdown to export the full lineage for audit workpapers.
    • Generates visual reports with charts, stat cards, and tables directly in the AI Assistant response — no file download required.
    • Supports side-by-side multi-chart dashboards in a single query (e.g., trend chart alongside a users-by-status chart), each with its own export option.
    • Composes full reports combining aggregate stat cards, trend charts, and data tables in one response.
    • Adds a View lineage Data lineage drawer showing the exact query and row count behind each report source, verification status against the assistant's execution log, and the full list of tool calls made to build the report.
    • Enables copying individual source queries or the entire lineage (sources, verification status, tool calls) as Markdown for audit workpapers.

    5 details in Everything

99. docs update Jul 28, 2026 · docs issue 004

    ConductorOne AI Assistant gains visual report generation with charts, stat cards, tables, and full data lineage verification.

    └──▷ TRY IT
    Render a side-by-side dashboard of grants trend and users by status directly in the AI Assistant.
    copy

    $ dashboard: grants trend and users by status side by side

    Audit a report's data provenanceprovenanceEvidence recording where a software artifact came from and how it was built, including its source and build process. It helps a consumer check that the artifact matches a trusted build. by opening the lineage drawer and exporting for workpapers.
    📍In the AI Assistant report, click View lineage to open the Data lineage drawer, review Verified/Unverified status and tool calls, then click Copy as Markdown to export the full lineage for audit workpapers.
    • Generates visual reports with charts, stat cards, and tables directly in the AI Assistant response — no file download required.
    • Supports side-by-side multi-chart dashboards in a single query (e.g., trend chart alongside a users-by-status chart), each with its own export option.
    • Composes full reports combining aggregate stat cards, trend charts, and data tables in one response.
    • Adds a View lineage Data lineage drawer showing the exact query and row count behind each report source, verification status against the assistant's execution log, and the full list of tool calls made to build the report.
    • Enables copying individual source queries or the entire lineage (sources, verification status, tool calls) as Markdown for audit workpapers.

    5 details in Everything

100. docs update Jul 28, 2026 · issue 028

     ConductorOne adds AI-driven workflow approvals, per-user connector auth, custom campaign recurrence, and richer AI reporting.

     • Assigns a workflow step in a request or review policy to C1 AI, which evaluates each request using built-in guidance plus custom instructions; every evaluation is logged as an AI-agent action in the audit log and recorded as a reviewable conversation.
     • Adds an indirect-access section to entitlement search results, surfacing entitlements (e.g., a GitHub team) that transitively grant access to a searched resource.
     • Enables custom recurrence intervals on campaign templates (e.g., every two days, every three months) and lets admins set exactly how long each campaign run lasts, replacing fixed presets.
     • Adds a button to open raw access-review and task-detail data as a readable, collapsible JSON tree on resource-level access reviews and in task details.
     • Supports adding groups to a request catalog or access profile so users can request group membership through the same approval flow as other requestable access.

     +5 more show less

     • Holds provisioning requests for entitlements not yet recognized by C1 (e.g., created via Terraform before a connector sync), shows the reason for the wait, resumes automatically after the next sync, and falls back to manual provisioning after 24 hours.
     • Supports combined step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) authentication and comment requirements on a single approval task, prompting for identity-provider verification and completing once verified.
     • Adds a View lineage action on C1 AI-generated reports to inspect the exact queries behind each number, verify each source against the execution log, and copy lineage for audit workpapers.
     • Renders multiple C1 AI-requested charts as an aligned grid with individual export and source links; C1 AI can also respond with a composed report combining stat cards, a trend chart, and a data table — each with sourcing.
     • Allows admins to require per-user connector credentials (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth) instead of a shared admin credential.

     10 details in Everything

101. docs update Jul 28, 2026 · docs issue 004

     ConductorOne adds AI-driven workflow approvals, per-user connector auth, custom campaign recurrence, and richer AI reporting.

     • Assigns a workflow step in a request or review policy to C1 AI, which evaluates each request using built-in guidance plus custom instructions; every evaluation is logged as an AI-agent action in the audit log and recorded as a reviewable conversation.
     • Adds an indirect-access section to entitlement search results, surfacing entitlements (e.g., a GitHub team) that transitively grant access to a searched resource.
     • Enables custom recurrence intervals on campaign templates (e.g., every two days, every three months) and lets admins set exactly how long each campaign run lasts, replacing fixed presets.
     • Adds a button to open raw access-review and task-detail data as a readable, collapsible JSON tree on resource-level access reviews and in task details.
     • Supports adding groups to a request catalog or access profile so users can request group membership through the same approval flow as other requestable access.

     +5 more show less

     • Holds provisioning requests for entitlements not yet recognized by C1 (e.g., created via Terraform before a connector sync), shows the reason for the wait, resumes automatically after the next sync, and falls back to manual provisioning after 24 hours.
     • Supports combined step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) authentication and comment requirements on a single approval task, prompting for identity-provider verification and completing once verified.
     • Adds a View lineage action on C1 AI-generated reports to inspect the exact queries behind each number, verify each source against the execution log, and copy lineage for audit workpapers.
     • Renders multiple C1 AI-requested charts as an aligned grid with individual export and source links; C1 AI can also respond with a composed report combining stat cards, a trend chart, and a data table — each with sourcing.
     • Allows admins to require per-user connector credentials (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth) instead of a shared admin credential.

     10 details in Everything

102. July 24, 2026 (2026-07-27) Jul 27, 2026 · issue 028

     ConductorOne adds AI agent policy steps, group/role access reviews, per-user MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. auth, and scoped personal API keys.

     └──▷ HOW TO FIND IT
     Delegate a policy step to the C1 AI agent so it automatically evaluates access requests and logs each decision for audit — no human reviewer needed for routine approvals.
     📍In the console, go to Policies › [your policy] › Edit, add or select a workflow step, choose Assign to agent, enter any custom instructions, and save. Audit log entries for that step will appear under the C1 System agent identity with the AI-agent action label.
     Scope a personal API key to read-only to limit blast radius when integrating with a third-party reporting tool.
     📍In the console, go to Settings › Personal API Keys › Create API Key, open the Limit Scopes selector, choose the read-only role, and generate the key.
     Run a quarterly access review that covers group and role grants, not just individual user grants, to satisfy broader compliance requirements.
     📍In the console, go to Access Reviews › Create Campaign, open the Scope tab, find the Grant types selector, choose Users and resources, set a custom recurrence interval (e.g., every three months), configure the campaign duration, and launch.
     • Assign a workflow step in a request or review policy to the C1 AI agent, which evaluates each request using built-in guidance plus custom instructions; every evaluation is attributed to the C1 System agent identity, labeled as an AI-agent action in the audit log, and recorded as a reviewable conversation.
     • Request catalog search now includes an Entitlements matching section showing entitlements that indirectly grant access to a searched resource (e.g., a GitHub team whose repo permissions include the target repo), with a View access graph link showing the full grant chain.
     • Access review campaigns can now review grants held by groups or roles in addition to users, using a new Grant types selector (Users only, Resources only, or Users and resources) on the Scope tab.
     • Campaign templates now support custom recurrence intervals (e.g., every two days, every three months) and configurable campaign durations, beyond the previous fixed presets.
     • IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy documents on AWS resources now render as a readable, collapsible JSON tree via a View policy button on resource-level access reviews and in task details.

     +10 more show less

     • C1 groups can now be added to a request catalog or access profile, making group membership requestable through the standard approval flow.
     • Requests for entitlements created ahead of a connector sync (e.g., via Terraform) now queue and wait rather than fail, display a status reason, resume automatically after the next sync, and fall back to manual provisioning after 24 hours.
     • A new Retry failed provisioning action on access tasks lets admins retry connector provisioning in place without canceling the task and losing collected approvals.
     • Step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) approvals that also require a comment now prompt for identity-provider verification in a single flow and complete once verified.
     • C1 AI assistant reports now include a View lineage button showing the exact queries behind each number, verified against the execution log, and copyable for audit workpapers.
     • C1 AI now renders multiple requested charts as an aligned grid with individual export and source links, and can respond with composed reports combining stat cards, a trend chart, and a data table.
     • A new Generate action in the CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. editor converts a plain-English description directly into a CEL condition expression in one step.
     • Admins can now require per-user authentication (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth) for external MCP servers instead of sharing a single admin credential.
     • Tool enablement settings (require enablement or auto-enable) can now be configured per MCP server, independently of the tenant-wide AI governance default.
     • Any user can now restrict a personal API key to a narrower role (e.g., read-only) using the Limit Scopes selector at key creation, without requiring Super Admin access.

     15 details in Everything

103. July 24, 2026 (2026-07-27) Jul 27, 2026 · issue 005

     ConductorOne adds AI agent policy steps, group/role access reviews, per-user MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. auth, and scoped personal API keys.

     └──▷ HOW TO FIND IT
     Delegate a policy step to the C1 AI agent so it automatically evaluates access requests and logs each decision for audit — no human reviewer needed for routine approvals.
     📍In the console, go to Policies › [your policy] › Edit, add or select a workflow step, choose Assign to agent, enter any custom instructions, and save. Audit log entries for that step will appear under the C1 System agent identity with the AI-agent action label.
     Scope a personal API key to read-only to limit blast radius when integrating with a third-party reporting tool.
     📍In the console, go to Settings › Personal API Keys › Create API Key, open the Limit Scopes selector, choose the read-only role, and generate the key.
     Run a quarterly access review that covers group and role grants, not just individual user grants, to satisfy broader compliance requirements.
     📍In the console, go to Access Reviews › Create Campaign, open the Scope tab, find the Grant types selector, choose Users and resources, set a custom recurrence interval (e.g., every three months), configure the campaign duration, and launch.
     • Assign a workflow step in a request or review policy to the C1 AI agent, which evaluates each request using built-in guidance plus custom instructions; every evaluation is attributed to the C1 System agent identity, labeled as an AI-agent action in the audit log, and recorded as a reviewable conversation.
     • Request catalog search now includes an Entitlements matching section showing entitlements that indirectly grant access to a searched resource (e.g., a GitHub team whose repo permissions include the target repo), with a View access graph link showing the full grant chain.
     • Access review campaigns can now review grants held by groups or roles in addition to users, using a new Grant types selector (Users only, Resources only, or Users and resources) on the Scope tab.
     • Campaign templates now support custom recurrence intervals (e.g., every two days, every three months) and configurable campaign durations, beyond the previous fixed presets.
     • IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy documents on AWS resources now render as a readable, collapsible JSON tree via a View policy button on resource-level access reviews and in task details.

     +10 more show less

     • C1 groups can now be added to a request catalog or access profile, making group membership requestable through the standard approval flow.
     • Requests for entitlements created ahead of a connector sync (e.g., via Terraform) now queue and wait rather than fail, display a status reason, resume automatically after the next sync, and fall back to manual provisioning after 24 hours.
     • A new Retry failed provisioning action on access tasks lets admins retry connector provisioning in place without canceling the task and losing collected approvals.
     • Step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) approvals that also require a comment now prompt for identity-provider verification in a single flow and complete once verified.
     • C1 AI assistant reports now include a View lineage button showing the exact queries behind each number, verified against the execution log, and copyable for audit workpapers.
     • C1 AI now renders multiple requested charts as an aligned grid with individual export and source links, and can respond with composed reports combining stat cards, a trend chart, and a data table.
     • A new Generate action in the CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. editor converts a plain-English description directly into a CEL condition expression in one step.
     • Admins can now require per-user authentication (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth) for external MCP servers instead of sharing a single admin credential.
     • Tool enablement settings (require enablement or auto-enable) can now be configured per MCP server, independently of the tenant-wide AI governance default.
     • Any user can now restrict a personal API key to a narrower role (e.g., read-only) using the Limit Scopes selector at key creation, without requiring Super Admin access.

     15 details in Everything

104. docs update Jul 25, 2026 · issue 025

     ConductorOne campaigns now support automated scheduling with configurable recurrence and duration.

     • Adds automated scheduling for campaign templates, enabling draft campaign instances to be created once on a future date or on a recurring schedule.
     • Supports configurable recurrence intervals (e.g., every 2 days, every 3 months) and campaign duration settings via the template's Automated scheduling section.

     2 details in Everything

105. docs update Jul 25, 2026 · docs issue 004

     ConductorOne campaigns now support automated scheduling with configurable recurrence and duration.

     • Adds automated scheduling for campaign templates, enabling draft campaign instances to be created once on a future date or on a recurring schedule.
     • Supports configurable recurrence intervals (e.g., every 2 days, every 3 months) and campaign duration settings via the template's Automated scheduling section.

     2 details in Everything

106. docs update Jul 25, 2026 · issue 025

     ConductorOne adds AI > Conversations audit page with per-tab visibility into agent activity across all three surfaces.

     └──▷ HOW TO FIND IT
     Review the reasoning and tool calls behind an AI agent's approval policy decision after an access request is evaluated.
     📍In the ConductorOne console, go to AI › Conversations and select the System tab to view policy step evaluations attributed to the C1 System agent.
     • New read-only AI > Conversations page lets admins review AI agent conversations across the organization, organized into separate tabs by surface.
     • New System tab surfaces agent policy step evaluations — including the agent's reasoning and tool calls — attributed to the built-in C1 System identity for every approval policy decision.

     2 details in Everything

107. docs update Jul 25, 2026 · docs issue 004

     ConductorOne adds AI > Conversations audit page with per-tab visibility into agent activity across all three surfaces.

     └──▷ HOW TO FIND IT
     Review the reasoning and tool calls behind an AI agent's approval policy decision after an access request is evaluated.
     📍In the ConductorOne console, go to AI › Conversations and select the System tab to view policy step evaluations attributed to the C1 System agent.
     • New read-only AI > Conversations page lets admins review AI agent conversations across the organization, organized into separate tabs by surface.
     • New System tab surfaces agent policy step evaluations — including the agent's reasoning and tool calls — attributed to the built-in C1 System identity for every approval policy decision.

     2 details in Everything

108. docs update Jul 25, 2026 · issue 025

     ConductorOne adds Retry failed provisioning to recover from connector failures without re-collecting approvals.

     └──▷ HOW TO FIND IT
     Recover a single access-change task stuck due to a broken connector group, preserving existing approvals.
     📍In the task detail view, open the menu and click Retry failed provisioning.
     Bulk-recover multiple tasks that failed provisioning at the same time (e.g., after a connector misconfiguration is fixed).
     📍In the task list, select the affected tasks, then choose Retry failed provisioning from the bulk actions menu.
     • Adds Retry failed provisioning action to re-attempt a failed connector provisioning step on an access-change task without restarting the approval workflow.
     • Supports bulk retry of multiple failed provisioning tasks via the bulk actions menu.

     2 details in Everything

109. docs update Jul 25, 2026 · docs issue 004

     ConductorOne adds Retry failed provisioning to recover from connector failures without re-collecting approvals.

     └──▷ HOW TO FIND IT
     Recover a single access-change task stuck due to a broken connector group, preserving existing approvals.
     📍In the task detail view, open the menu and click Retry failed provisioning.
     Bulk-recover multiple tasks that failed provisioning at the same time (e.g., after a connector misconfiguration is fixed).
     📍In the task list, select the affected tasks, then choose Retry failed provisioning from the bulk actions menu.
     • Adds Retry failed provisioning action to re-attempt a failed connector provisioning step on an access-change task without restarting the approval workflow.
     • Supports bulk retry of multiple failed provisioning tasks via the bulk actions menu.

     2 details in Everything

110. docs update Jul 25, 2026 · issue 025

     ConductorOne now renders attached resource attributes (e.g. AWS IAMAWS IAMAWS Identity and Access Management is Amazon's service for controlling who and what can call AWS APIs, by attaching policies that allow or deny actions on resources. Cyber tools use it to scope credentials to least privilege. policy documents) as a collapsible JSON tree in the review task details drawer.

     └──▷ HOW TO FIND IT
     Inspect an AWS IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy document attached to a resource during an access review without wading through raw JSON.
     📍In the review task details drawer, locate the resource attribute that contains a policy document, then click View policy to expand it as a collapsible JSON tree.
111. docs update Jul 25, 2026 · docs issue 004

     ConductorOne now renders attached resource attributes (e.g. AWS IAMAWS IAMAWS Identity and Access Management is Amazon's service for controlling who and what can call AWS APIs, by attaching policies that allow or deny actions on resources. Cyber tools use it to scope credentials to least privilege. policy documents) as a collapsible JSON tree in the review task details drawer.

     └──▷ HOW TO FIND IT
     Inspect an AWS IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy document attached to a resource during an access review without wading through raw JSON.
     📍In the review task details drawer, locate the resource attribute that contains a policy document, then click View policy to expand it as a collapsible JSON tree.
112. docs update Jul 25, 2026 · issue 025

     ConductorOne adds AI agent policy steps, group/role access reviews, custom campaign schedules, and per-user MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. auth

     └──▷ HOW TO FIND IT
     Set up an AI agent step in an approval policy to auto-evaluate requests with custom criteria, so routine low-risk requests are handled without human reviewers.
     📍In the console, go to Policies › [your policy] › Add Step › Agent. Set Agent mode to Full control, Agent failure action to Reassign to super admins, and enter your approval criteria in Agent instructions. Save the policy.
     Restrict a personal API key to read-only access when sharing it with a monitoring tool, regardless of your own role.
     📍In the console, go to your profile › API Keys › Create API Key. In the Limit Scopes selector, choose Read-only. Complete key creation and copy the key.
     Find the right entitlement to request when you only know the resource name (e.g., a GitHub repo), not which group owns it.
     📍In the request catalog search bar, type the resource name (e.g., analytics-backend). Under Entitlements matching, locate the relevant GitHub team entry and click View access graph to confirm the grant chain before requesting.
     • Assign an AI Agent step to any request or review policy workflow, configuring agent mode, failure action, and free-text approval/deny instructions; every evaluation is attributed to the C1 System identity and logged under AI > Conversations.
     • Catalog search now returns an Entitlements matching section showing entitlements that indirectly grant access to the searched resource, with a View access graph link to see the full grant chain.
     • Access review campaigns can now review grants held by groups or roles (not just users) via a new Grant Types selector offering Users only, Resources only, or Users and resources.
     • Campaign templates support custom recurrence intervals (e.g., every 2 days or every 3 months) and configurable campaign duration, beyond the previous fixed presets.
     • IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy documents on AWS resources now render as a readable, collapsible JSON tree via a View policy button, available in resource-level access reviews and task details.

     +12 more show less

     • Requests for entitlements created ahead of a connector sync (e.g., via Terraform) now wait and resume automatically after sync completes, with a 24-hour manual-provisioning fallback.
     • C1 groups can now be added to a request catalog or access profile, making group membership requestable through the standard approval, duration, and revocation flow.
     • Failed connector provisioning on an access task can now be retried directly via Retry failed provisioning without cancelling the task or losing collected approvals.
     • Step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) approvals that also require a comment can now be completed together in a single flow.
     • External MCP servers now support per-user authentication (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth) instead of requiring a shared admin credential.
     • AI-generated reports include a Data lineage view showing the exact queries behind each number, verified against the execution log, with copy support for audit workpapers.
     • Ask C1 AI now renders multiple requested charts as an aligned grid and can produce composed reports combining stat cards, a trend chart, and a data table in one response.
     • The CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. editor gains a Generate action that converts a plain-English description directly into a CEL expression in one step.
     • Tool enablement (require enablement vs. auto-enable) can now be set per MCP server, independent of the tenant-level AI governance default.
     • External and hosted MCP servers that fail relaxable OAuth discovery checks (e.g., issuer mismatch) can now be registered after an admin reviews and accepts the flagged findings.
     • Hosted MCP servers now default to using the tool prefix declared by the connector itself rather than a generic derived name.
     • Any user can now use the Limit Scopes selector when creating a personal API key to restrict it to a narrower role (e.g., read-only), without requiring Super Admin access.

     17 details in Everything

113. docs update Jul 25, 2026 · docs issue 004

     ConductorOne adds AI agent policy steps, group/role access reviews, custom campaign schedules, and per-user MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. auth

     └──▷ HOW TO FIND IT
     Set up an AI agent step in an approval policy to auto-evaluate requests with custom criteria, so routine low-risk requests are handled without human reviewers.
     📍In the console, go to Policies › [your policy] › Add Step › Agent. Set Agent mode to Full control, Agent failure action to Reassign to super admins, and enter your approval criteria in Agent instructions. Save the policy.
     Restrict a personal API key to read-only access when sharing it with a monitoring tool, regardless of your own role.
     📍In the console, go to your profile › API Keys › Create API Key. In the Limit Scopes selector, choose Read-only. Complete key creation and copy the key.
     Find the right entitlement to request when you only know the resource name (e.g., a GitHub repo), not which group owns it.
     📍In the request catalog search bar, type the resource name (e.g., analytics-backend). Under Entitlements matching, locate the relevant GitHub team entry and click View access graph to confirm the grant chain before requesting.
     • Assign an AI Agent step to any request or review policy workflow, configuring agent mode, failure action, and free-text approval/deny instructions; every evaluation is attributed to the C1 System identity and logged under AI > Conversations.
     • Catalog search now returns an Entitlements matching section showing entitlements that indirectly grant access to the searched resource, with a View access graph link to see the full grant chain.
     • Access review campaigns can now review grants held by groups or roles (not just users) via a new Grant Types selector offering Users only, Resources only, or Users and resources.
     • Campaign templates support custom recurrence intervals (e.g., every 2 days or every 3 months) and configurable campaign duration, beyond the previous fixed presets.
     • IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy documents on AWS resources now render as a readable, collapsible JSON tree via a View policy button, available in resource-level access reviews and task details.

     +12 more show less

     • Requests for entitlements created ahead of a connector sync (e.g., via Terraform) now wait and resume automatically after sync completes, with a 24-hour manual-provisioning fallback.
     • C1 groups can now be added to a request catalog or access profile, making group membership requestable through the standard approval, duration, and revocation flow.
     • Failed connector provisioning on an access task can now be retried directly via Retry failed provisioning without cancelling the task or losing collected approvals.
     • Step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) approvals that also require a comment can now be completed together in a single flow.
     • External MCP servers now support per-user authentication (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth) instead of requiring a shared admin credential.
     • AI-generated reports include a Data lineage view showing the exact queries behind each number, verified against the execution log, with copy support for audit workpapers.
     • Ask C1 AI now renders multiple requested charts as an aligned grid and can produce composed reports combining stat cards, a trend chart, and a data table in one response.
     • The CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. editor gains a Generate action that converts a plain-English description directly into a CEL expression in one step.
     • Tool enablement (require enablement vs. auto-enable) can now be set per MCP server, independent of the tenant-level AI governance default.
     • External and hosted MCP servers that fail relaxable OAuth discovery checks (e.g., issuer mismatch) can now be registered after an admin reviews and accepts the flagged findings.
     • Hosted MCP servers now default to using the tool prefix declared by the connector itself rather than a generic derived name.
     • Any user can now use the Limit Scopes selector when creating a personal API key to restrict it to a narrower role (e.g., read-only), without requiring Super Admin access.

     17 details in Everything

114. July 24, 2026 (2026-07-25) Jul 25, 2026 · issue 025

     ConductorOne adds AI agent policy steps, access-graph search, group/role reviews, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. per-user auth, and scoped API keys.

     └──▷ HOW TO FIND IT
     Set up an AI agent step in a policy to auto-evaluate low-risk access requests and reassign to super admins if it can't decide.
     📍In the console, go to Policies › [your policy] › Edit, add a new step, select Agent as the step type, set Agent mode to Full control, set Agent failure action to Reassign to super admins, and enter Agent instructions such as 'Approve if the requester is in the contractors group and the resource is tagged low-risk; otherwise deny.'
     Scope a personal API key to read-only to safely hand credentials to a reporting script without granting full account access.
     📍In the console, go to your profile › API Keys › Create API Key, then open the Limit Scopes selector and choose the read-only role before saving.
     Include groups and roles in a quarterly access review campaign to catch over-provisioned shared identities alongside user grants.
     📍In the console, go to Access Reviews › Campaigns › Create Campaign, open the Scope tab, and set the Grant types selector to Users and resources before launching.
     • Adds AI Agent steps to request and review policies — configure agent mode (comment-only, full control, or policy-step changes), failure action (reassign or skip), and free-text approval instructions; evaluations appear in audit log and AI > Conversations.
     • Extends request catalog search to surface entitlements that indirectly grant access to a searched resource, with a View access graph link showing the full grant chain.
     • Access review campaigns can now scope to groups and roles (not just users) via a new Grant types selector with Users only, Resources only, or Users and resources options.
     • Campaign templates support custom recurrence intervals (e.g. every 2 days or every 3 months) and configurable campaign run duration.
     • IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy documents on AWS resources now render as a readable, collapsible JSON tree via a View policy button in resource-level access reviews and task details.

     +12 more show less

     • Requests for entitlements created ahead of connector sync (e.g. via Terraform) now queue and auto-resume after sync, instead of failing immediately.
     • C1 groups can now be added to a request catalog or access profile, making group membership requestable through standard approval, duration, and revocation flows.
     • Adds Retry failed provisioning action on access tasks, allowing provisioning retries without canceling or losing collected approvals.
     • Step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) approvals that also require a comment can now be completed together in a single flow.
     • Per-user authentication for external MCP servers now supports OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth — replacing shared admin credentials.
     • New Data lineage view for AI-generated reports shows exact queries behind each number and lets users copy lineage for audit workpapers.
     • C1 AI now renders multiple requested charts as an aligned grid and can produce composed reports combining stat cards, trend charts, and data tables in one response.
     • New Generate action in the CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. editor converts plain-English descriptions directly into CEL expressions without requiring iterative AI chat.
     • Tool enablement (require enablement vs. auto-enable) can now be set per MCP server, independent of the tenant-wide AI governance default.
     • Admins can now register MCP servers that fail relaxable OAuth discovery checks (e.g. issuer mismatch, cross-host authorization server) by reviewing and acknowledging flagged findings.
     • Hosted MCP servers without an admin-set tool prefix now default to the prefix declared by the connector itself, keeping tool names recognizable to AI clients.
     • Any user can now use the Limit Scopes selector when creating a personal API key to restrict it to a narrower role (e.g. read-only), without requiring Super Admin access.

     17 details in Everything

115. July 24, 2026 (2026-07-25) Jul 25, 2026 · issue 005

     ConductorOne adds AI agent policy steps, access-graph search, group/role reviews, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. per-user auth, and scoped API keys.

     └──▷ HOW TO FIND IT
     Set up an AI agent step in a policy to auto-evaluate low-risk access requests and reassign to super admins if it can't decide.
     📍In the console, go to Policies › [your policy] › Edit, add a new step, select Agent as the step type, set Agent mode to Full control, set Agent failure action to Reassign to super admins, and enter Agent instructions such as 'Approve if the requester is in the contractors group and the resource is tagged low-risk; otherwise deny.'
     Scope a personal API key to read-only to safely hand credentials to a reporting script without granting full account access.
     📍In the console, go to your profile › API Keys › Create API Key, then open the Limit Scopes selector and choose the read-only role before saving.
     Include groups and roles in a quarterly access review campaign to catch over-provisioned shared identities alongside user grants.
     📍In the console, go to Access Reviews › Campaigns › Create Campaign, open the Scope tab, and set the Grant types selector to Users and resources before launching.
     • Adds AI Agent steps to request and review policies — configure agent mode (comment-only, full control, or policy-step changes), failure action (reassign or skip), and free-text approval instructions; evaluations appear in audit log and AI > Conversations.
     • Extends request catalog search to surface entitlements that indirectly grant access to a searched resource, with a View access graph link showing the full grant chain.
     • Access review campaigns can now scope to groups and roles (not just users) via a new Grant types selector with Users only, Resources only, or Users and resources options.
     • Campaign templates support custom recurrence intervals (e.g. every 2 days or every 3 months) and configurable campaign run duration.
     • IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. policy documents on AWS resources now render as a readable, collapsible JSON tree via a View policy button in resource-level access reviews and task details.

     +12 more show less

     • Requests for entitlements created ahead of connector sync (e.g. via Terraform) now queue and auto-resume after sync, instead of failing immediately.
     • C1 groups can now be added to a request catalog or access profile, making group membership requestable through standard approval, duration, and revocation flows.
     • Adds Retry failed provisioning action on access tasks, allowing provisioning retries without canceling or losing collected approvals.
     • Step-up (MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone.) approvals that also require a comment can now be completed together in a single flow.
     • Per-user authentication for external MCP servers now supports OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth — replacing shared admin credentials.
     • New Data lineage view for AI-generated reports shows exact queries behind each number and lets users copy lineage for audit workpapers.
     • C1 AI now renders multiple requested charts as an aligned grid and can produce composed reports combining stat cards, trend charts, and data tables in one response.
     • New Generate action in the CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. editor converts plain-English descriptions directly into CEL expressions without requiring iterative AI chat.
     • Tool enablement (require enablement vs. auto-enable) can now be set per MCP server, independent of the tenant-wide AI governance default.
     • Admins can now register MCP servers that fail relaxable OAuth discovery checks (e.g. issuer mismatch, cross-host authorization server) by reviewing and acknowledging flagged findings.
     • Hosted MCP servers without an admin-set tool prefix now default to the prefix declared by the connector itself, keeping tool names recognizable to AI clients.
     • Any user can now use the Limit Scopes selector when creating a personal API key to restrict it to a narrower role (e.g. read-only), without requiring Super Admin access.

     17 details in Everything

116. docs update Jul 24, 2026 · issue 024

     ConductorOne campaigns can now review access grants held by groups and roles, not just users.

     └──▷ HOW TO FIND IT
     Run an access review that covers group and role grants — useful for auditing privileged resource-level access separate from individual user reviews.
     📍In the campaign (or campaign template) editor, under Add attribute, select Grant type, then click Change selection and choose Resources only to review grants held by groups or roles.
     • Adds a Grant type attribute to campaigns and campaign templates, enabling access reviews scoped to users only, resources only (groups/roles), or both users and resources.

     1 detail in Everything

117. docs update Jul 24, 2026 · docs issue 004

     ConductorOne campaigns can now review access grants held by groups and roles, not just users.

     └──▷ HOW TO FIND IT
     Run an access review that covers group and role grants — useful for auditing privileged resource-level access separate from individual user reviews.
     📍In the campaign (or campaign template) editor, under Add attribute, select Grant type, then click Change selection and choose Resources only to review grants held by groups or roles.
     • Adds a Grant type attribute to campaigns and campaign templates, enabling access reviews scoped to users only, resources only (groups/roles), or both users and resources.

     1 detail in Everything

118. docs update Jul 24, 2026 · issue 024

     ConductorOne MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers now support JWTJWTAn open standard (RFC 7519) for encoding claims as a signed, optionally encrypted JSON object passed between parties. Cyber tools use it to verify identity and permissions without querying a central session store. Bearer (RFC 7523RFC 7523An IETF standard that defines how a JSON Web Token (JWT) can be used as a credential to request an OAuth 2.0 access token, letting a cyber tool authenticate to an API without exchanging a username and password at runtime.) authentication.

119. docs update Jul 24, 2026 · docs issue 004

     ConductorOne MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers now support JWTJWTAn open standard (RFC 7519) for encoding claims as a signed, optionally encrypted JSON object passed between parties. Cyber tools use it to verify identity and permissions without querying a central session store. Bearer (RFC 7523RFC 7523An IETF standard that defines how a JSON Web Token (JWT) can be used as a credential to request an OAuth 2.0 access token, letting a cyber tool authenticate to an API without exchanging a username and password at runtime.) authentication.

120. docs update Jul 24, 2026 · issue 024

     ConductorOne adds MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. client support and controls to disable MCP traffic and system features

     └──▷ HOW TO FIND IT
     When you want to prevent MCP-based AI client access to your environment, disable MCP traffic from the admin settings.
     📍In the console, go to Admin › Enable AI Access Management, then toggle Disable MCP traffic.
     • Supports MCP (Model Context ProtocolModel Context ProtocolModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface.) clients for AI access management integration.
     • Adds admin controls to disable MCP traffic.
     • Adds admin controls to disable system features.

     3 details in Everything

121. docs update Jul 24, 2026 · docs issue 004

     ConductorOne adds MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. client support and controls to disable MCP traffic and system features

     └──▷ HOW TO FIND IT
     When you want to prevent MCP-based AI client access to your environment, disable MCP traffic from the admin settings.
     📍In the console, go to Admin › Enable AI Access Management, then toggle Disable MCP traffic.
     • Supports MCP (Model Context ProtocolModel Context ProtocolModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface.) clients for AI access management integration.
     • Adds admin controls to disable MCP traffic.
     • Adds admin controls to disable system features.

     3 details in Everything

122. docs update Jul 24, 2026 · issue 024

     ConductorOne AI assistant now accessible via Slack DMs in addition to the web app

     └──▷ HOW TO FIND IT
     Query C1 identity data from an external AI client without using the web app.
     📍Use the C1 MCP integration to connect external AI clients such as Claude Desktop or Cursor to C1 identity data. See the C1 MCP integration docs for setup steps.
     • Enables interaction with the C1 AI assistant directly via Slack DMs once the C1 app for Slack is installed.
     • Documents the C1 MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. integration as a programmatic path for external AI clients (such as Claude Desktop or Cursor) to query C1 identity data.

     2 details in Everything

123. docs update Jul 24, 2026 · docs issue 004

     ConductorOne AI assistant now accessible via Slack DMs in addition to the web app

     └──▷ HOW TO FIND IT
     Query C1 identity data from an external AI client without using the web app.
     📍Use the C1 MCP integration to connect external AI clients such as Claude Desktop or Cursor to C1 identity data. See the C1 MCP integration docs for setup steps.
     • Enables interaction with the C1 AI assistant directly via Slack DMs once the C1 app for Slack is installed.
     • Documents the C1 MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. integration as a programmatic path for external AI clients (such as Claude Desktop or Cursor) to query C1 identity data.

     2 details in Everything

124. Product digest: July 24, 2026 Jul 24, 2026 · release notes

     ConductorOne adds C1 AI policy step agents, direct task retry, CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. generator, scoped personal API keys, and expanded search

     └──▷ HOW TO FIND IT
     Retry provisioning on a failed access task without losing existing approvals.
     📍Open the task details -> Click Retry failed provisioning
     Create a personal API key with restricted scopes below your own role.
     📍In Create Personal API Key UI, use the Limit Scopes selector to choose a narrower role (e.g., read-only).
     Configure an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server to require per-user authentication instead of a shared admin credential.
     📍In MCP server settings, enable Per-user authentication and select OAuth2, bearer token, custom header, or basic auth.
     • C1 AI agent assignment to request/review policy steps for automated evaluation with audit attribution
     • Entitlements matching in search results that reveals indirect access chains via View access graph
     • Grant types selector in access review campaigns scope tab enabling reviews of grants held by groups or roles
     • View policy button on resource attributes rendering AWS IAMAWS IAMAWS Identity and Access Management is Amazon's service for controlling who and what can call AWS APIs, by attaching policies that allow or deny actions on resources. Cyber tools use it to scope credentials to least privilege. policies as a collapsible JSON tree
     • C1 groups added to request catalog and access profiles for group membership requests

     +8 more show less

     • Retry failed provisioning action on access tasks allowing retry without restarting the task
     • Auto step-up authentication prompt combined with comment requirement in task approvals
     • Data lineage view for C1 AI agentic reports showing exact queries and verification status
     • Multi-chart dashboards and composed reports in C1 AI assistant responses
     • Generate action in CEL editor producing conditions from plain English input
     • Per-user authentication options for external MCP servers (OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, basic auth)
     • Tool enablement settings configurable on individual MCP servers independent of tenant default
     • Limit Scopes selector when creating personal API keys to restrict permissions below user role

     13 details in Everything

125. July 10, 2026 (2026-07-10) Jul 10, 2026 · issue 014

     ConductorOne adds real-time access provisioning, catalog indirect-access search, bulk revoke, visual condition builder, and Terraform connector ownership resources.

     └──▷ USE IT
     Migrate connector ownership in Terraform to avoid drift after the user_ids field is dropped from conductorone_integration_* resources.
     hcl copy

     resource "conductorone_connector_owner_user" "example" {
       connector_id = conductorone_integration_aws.example.connector_id
       user_id      = "<user-id>"
     }
     
     resource "conductorone_connector_owner_entitlement" "example" {
       connector_id   = conductorone_integration_aws.example.connector_id
       entitlement_id = "<entitlement-id>"
     }

     Connect an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server without managing client credentials when the server supports OAuth Dynamic Client Registration.
     📍In the C1 console, go to AI › MCP, add a new hosted or external MCP server, and enable Use dynamic client registration — C1 will handle OAuth registration automatically with no client ID or secret required.
     Filter which Directory source accounts are imported into C1 using the visual condition builder with a live preview before saving.
     📍In the C1 console, go to Directory sources settings, open the three-dot menu next to your directory, select Condition, build your import filter visually, then click Preview to verify the match count and spot-check individual users before saving.
     • Adds real-time streaming of access profile grants and removals to connected apps, replacing hourly sync cycles.
     • New rolling-hour anomaly detector for membership removals, with a viewable and adjustable threshold in automation settings.
     • Request catalog now surfaces the role, team, or group that grants access to a searched resource, even when the entitlement name doesn't match the search term.
     • Enables bulk revoke of entitlements directly from the App Accounts page — select one or more accounts and remove specific entitlements or all access at once.
     • New visual condition builder for Directory source import filters replaces raw CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expressions, with a built-in Data Preview to verify match counts before saving.

     +15 more show less

     • New top-level navigation sections: AI (with MCP server config and C1 Gateway) and Identity security (with Inventory).
     • C1 AI assistant in Slack now sends rich interactive responses — buttons, forms, and pickers — enabling task completion without leaving Slack.
     • Agentic AI Conversations page adds separate Web and Slack tabs for full cross-surface agent activity visibility.
     • MCP servers supporting OAuth Dynamic Client Registration can now be connected without entering a client ID or secret.
     • Adds per-user authentication for external MCP servers, supporting OAuth2OAuth2An open authorization framework (RFC 6749) that lets a service grant limited access to a user's resources on another service without exposing credentials, giving cyber tools a standard way to authenticate against third-party APIs., bearer tokenbearer tokenA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else., custom header, or basic auth instead of a shared admin credential.
     • Toolset contents are now visible on hover in the task view, without navigating to entitlement detail pages.
     • Conflict monitors now support an Is not in condition for Group B, enabling detection of sensitive access held without a required compensating entitlement.
     • Column configurator extended to 23 additional pages — including Findings, Automations, Policies, MCP Servers, Secrets, and Webhooks — with preferences saved across sessions.
     • Entitlements table and CSV export now include an Owners column per application.
     • Row-by-row drawer navigation adds next/previous controls so users can move through list-view rows without reopening drawers.
     • Markdown (links, bold, lists) now supported in requestable form field helper text.
     • New System management setting lets admins disable connector anomaly pauses across all connectors tenant-wide.
     • System Owners now see service principal credentials expiring within 30 days (or already expired) in the daily digest email, with a countdown and direct link.
     • Admins can now edit post actions on the built-in App owner certification policy directly in the UI, without using the API.
     • Breaking change: Terraform provider v1.4.0 introduces two new resources — conductorone_connector_owner_user and conductorone_connector_owner_entitlement — to replace the deprecated user_ids field on all conductorone_integration_* resources; existing user_ids configs will cause drift.

     20 details in Everything

     └──▷ BREAKING ON UPGRADE

     • !The user_ids field on all conductorone_integration_* Terraform resources will no longer be returned by the Terraform provider starting in provider v1.4.0; if user_ids is configured in your Terraform files, this will cause drift and requires migration to the new conductorone_connector_owner_user and conductorone_connector_owner_entitlement resources and removal of the user_ids field.
     • !In the C1 navigation, User data sources in the Directory section is renamed to Directory sources and Groups is renamed to Dynamic groups; Inventory has moved from Explore to the new top-level Identity security section; MCP server configuration and C1 Gateway settings have moved under the new top-level AI section.
126. Product digest: July 10, 2026 Jul 10, 2026 · release notes

     Real-time profile sync, visual dir condition builder, AI Gov Admin role, and DCR for MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers

     └──▷ HOW TO FIND IT
     Connect a hosted MCP server using dynamic registration instead of manual client credentials.
     📍Enable Use dynamic client registration when adding the server in the C1 UI.
     Delegate AI access governance to platform team members without full admin rights.
     📍Assign the AI Governance Administrator user role to dedicated security or platform team members.
     Migrate Terraform connector ownership to avoid drift after upgrading provider v1.4.0.
     terraform copy

     Remove `user_ids` from `conductorone_integration_*` resources and use `conductorone_connector_owner_user` / `conductorone_connector_owner_entitlement` instead.

     • Adds real-time streaming for access profile grants and removals to connected apps instead of hourly syncs.
     • Adds a threshold setting in the access profile membership automation settings to configure anomaly detection for member removals.
     • Adds bulk revoke access capabilities on the app Accounts page and Grants tab for app owners.
     • Replaces raw CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. input with a visual condition builder for Directory sources, including a built-in Data preview.
     • Promotes AI to a top-level navigation section containing MCP server configuration (AI > MCP) and C1 Gateway settings (AI > C1 Gateway).

     +12 more show less

     • Creates a new AI Governance Administrator user role for delegating AIAM governance without Super Admin privileges.
     • Enables Slack interactions with rich interactive responses including buttons, forms, and pickers.
     • Adds separate Web and Slack tabs on the Agentic AI Conversations page in the admin console for audit visibility.
     • Adds support for OAuth Dynamic Client Registration (DCR) to connect hosted or external MCP servers without client credentials.
     • Allows hovering over toolsets in the task view to see included tools directly.
     • Adds an Is not in condition type for Group B in conflict monitors.
     • Extends column configurator availability to 23 additional pages including Findings, Automations, and MCP Servers.
     • Adds an Entitlement Owners column to the application Entitlements table and CSV exports.
     • Adds next/previous controls to drawers opened from list views for row-by-row navigation.
     • Supports markdown in requestable form field helper text.
     • Adds a System management setting to disable connector anomaly pauses tenant-wide.
     • Includes expiring service principal credentials in the Super Administrator daily digest email.

     17 details in Everything

     └──▷ BREAKING ON UPGRADE

     • !The user_ids field on all conductorone_integration_* resources will no longer be returned by the Terraform provider, causing drift if configured; ownership is now managed via conductorone_connector_owner_user and conductorone_connector_owner_entitlement resources in provider v1.4.0 or later.
127. Private bridge for MCP servers (2026-06-26) Jun 26, 2026 · issue 014

     ConductorOne now governs private/air-gapped MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers via an outbound bridge agent — no inbound firewall rules required.

     └──▷ HOW TO FIND IT
     Deploy the bridge agent next to a private MCP server so its tools are governed by C1 without exposing inbound network access.
     📍In the C1 console, ensure AI access management is enabled for your tenant, then follow Settings › Connect a private MCP server through a bridge to deploy the bridge-client agent alongside your private MCP server.
     • Adds private MCP server connectivity via a lightweight bridge-client agent that maintains an outbound connection to C1, eliminating the need to open inbound firewall rules or expose servers to the public internet.
     • Enables full C1 entitlement governance (access requests, approvals, reviews, and auditing) for MCP tools running on-prem, in air-gapped clusters, or behind firewalls.

     2 details in Everything

128. Private bridge for MCP servers (2026-06-26) Jun 26, 2026 · release notes

     Adds bridge-client agent and private MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server governance via outbound connection

     • Introduces the bridge-client lightweight bridge agent to connect on-prem or air-gapped MCP servers to ConductorOne without inbound firewall rules.
     • Enables governing tools for private MCP servers through C1 with standard entitlement requests, approvals, reviews, and audits.

     2 details in Everything

129. June 26, 2026 (2026-06-26) Jun 26, 2026 · issue 014

     ConductorOne adds single-page role mining, custom email providers, bulk role assignment, and flexible table exports.

     └──▷ HOW TO FIND IT
     Surface all denied or errored tasks in a large campaign without paging through the full task list.
     📍In the console, go to Campaigns › [your campaign] › Tasks, then open the Outcome filter and select Denied or Error to isolate tasks that need follow-up.
     Onboard a new team by assigning the same C1 role to many users at once instead of editing each profile individually.
     📍In the console, go to Settings › Users, select the checkboxes for each new team member, and choose the bulk Change role action to assign their role in one step.
     Produce a clean, auditor-ready policy export containing only the columns your compliance team needs.
     📍In the console, navigate to a policy's detail page or the policy list page and click Export to CSV to download the policy rules directly.
     • Redesigns role mining as a single-page experience with a token-based filter bar, inline cohort refinements, and a floating action bar for creating or updating access profiles without navigating away.
     • Adds early-access custom email provider support (Google Workspace, AWS SESAWS SESAmazon's cloud email-sending service, used by tools that need to deliver alerts, reports, or notifications without running their own mail infrastructure., Microsoft 365, SendGrid) so C1 notifications send from your own domain instead of no-reply@conductorone.com.
     • Enables bulk role assignment — select multiple users on the Users page and apply a role change to all of them at once.
     • Introduces customizable, reorderable columns (saved per table) and targeted CSV export for the Apps, App accounts, Entitlements, and Resources tables.
     • Adds outcome-based filtering on campaign task lists (granted, denied, revoked, certified, error, canceled, timed out) to quickly surface tasks needing attention.

     +7 more show less

     • Publishes step-by-step setup guides for 38 catalog MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers, covering credential creation, authentication, and tool discovery.
     • Adds an Access was inappropriate pre-written reason to access removal, revocation, and certification comment dialogs.
     • Shows the complete inbound webhook URL in the trigger drawer, ready to copy, eliminating manual URL construction from a listener ID.
     • Introduces a Current step filter on the task log to surface tasks awaiting approval or awaiting provisioning directly.
     • Enables sorting of the MCP server Tools table by tool name, visibility, classification, state, and last updated.
     • Adds CSV export of policy rules directly from a policy's detail page or the policy list page for easier auditing.
     • Replaces inline owner entry when creating apps with the full Select owners modal, including search, filtering, and Users/Entitlements tabs.

     12 details in Everything

130. June 26, 2026 (2026-06-26) Jun 26, 2026 · release notes

     Role mining refresh, custom email provider early access, bulk role assignment, CSV exports, and campaign task filtering

     • Adds Role mining interface with token-based filter bar, inline recommendations, and floating action bar for refined cohort analysis.
     • Early access to Custom email provider allowing C1 notifications to be sent from your own domain (Google Workspace, AWS SESAWS SESAmazon's cloud email-sending service, used by tools that need to deliver alerts, reports, or notifications without running their own mail infrastructure., Microsoft 365, SendGrid).
     • Adds Bulk role assignment capability via the Change role action on the Users page to select and assign roles to multiple users at once.
     • Enables customizable columns and CSV export for Apps, App accounts, Entitlements, and Resources tables via the Download to CSV drawer.
     • Adds filtering for campaign task lists by outcome (granted, denied, revoked, certified, error, canceled, timed out).

     +2 more show less

     • Provides setup guides for 38 catalog MCP servers covering credential creation, authentication, and tool discovery.
     • Adds Export policy rules to CSV directly from a policy's detail page or the policy list page.

     7 details in Everything

131. Product digest: June 26, 2026 Jun 26, 2026 · release notes

     ConductorOne adds role mining UI refresh, custom email providers, bulk role assignment, CSV exports for tables and policy rules, and task filters.

     └──▷ HOW TO FIND IT
     Assign roles to multiple users simultaneously on the Users page.
     📍Select multiple users › Bulk Change role
     Filter campaign tasks by specific outcome to surface issues.
     📍Campaign Task List › Filter by outcome: denied
     • Adds custom email provider setting to send C1 notifications from your own domain (Google Workspace, AWS SESAWS SESAmazon's cloud email-sending service, used by tools that need to deliver alerts, reports, or notifications without running their own mail infrastructure., Microsoft 365, SendGrid)
     • Adds Bulk Change role action on the Users page for assigning roles to multiple users at once
     • Adds CSV export via the Download to CSV drawer with customizable column selection for Apps, App accounts, Entitlements, and Resources tables
     • Adds filtering of campaign task lists by outcome (granted, denied, revoked, certified, error, canceled, timed out)
     • Adds a Current step filter in the Task log to find tasks awaiting approval or provisioning

     +1 more show less

     • Enables export of policy rules to CSV from a policy's detail page or the policy list page

     6 details in Everything

132. Ask C1AI (2026-06-24) Jun 24, 2026 · issue 014

     ConductorOne launches Ask C1AI: a floating AI chat assistant for querying access data and taking action from anywhere in the platform.

     └──▷ HOW TO FIND IT
     While reviewing a policy page, open Ask C1AI to interrogate entitlements in context and act on findings without leaving the page.
     📍In C1, navigate to any page (e.g. an application or policy view) › click Ask C1AI in the bottom-right corner › use the context toggle to include the current page context › type your question about access or entitlements.
     • New Ask C1AI floating chat panel available on every page lets practitioners query org access data and take action without navigating away from their current view.
     • AI assistant ingests page context — the application, entitlements, or policy currently in view — to deliver more precise answers, with a toggle to include or exclude that context per conversation.

     2 details in Everything

133. Ask C1AI (2026-06-24) Jun 24, 2026 · release notes

     Ask C1AI floating chat panel launched to query org access data without leaving the page

     └──▷ HOW TO FIND IT
     Query entitlements or policies while reviewing them in the UI without navigating to a separate report.
     📍Click Ask C1AI in the bottom-right corner of any page to open the chat panel.
134. Enterprise-managed authorization (2026-06-18) Jun 18, 2026 · issue 014

     ConductorOne adds enterprise-managed authorization for AI agents, issuing short-lived scoped tokens via the open XAA standard.

     • New early-access Enterprise-managed authorization enables AI agents (Claude, VS Code, etc.) to access organization-run MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers using short-lived, scoped tokens — no per-tool prompts, no persistent secrets.
     • Agents receive tokens only after entitlement checks, with every token request recorded for auditability.
     • AI agent access follows the same request, approval, review, and revocation workflows used for all other access in ConductorOne.
     • Built on the open Cross-App Access (XAA) standard for interoperability.

     4 details in Everything

135. Enterprise-managed authorization (2026-06-18) Jun 18, 2026 · release notes

     Early access to Enterprise-managed authorization for AI agents via XAA standard

     • Adds Enterprise-managed authorization feature enabling AI agents (e.g., Claude, VS Code) to reach MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers using short-lived scoped tokens instead of per-tool auth prompts or persistent secrets

     1 detail in Everything

136. June 12, 2026 (2026-06-12) Jun 12, 2026 · release notes

     ConductorOne adds Slack AI assistant for actions, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server verification, annotation-based routing, customizable campaign columns, and multi-recipient credential delivery.

     └──▷ HOW TO FIND IT
     Test the configuration of an external MCP server before saving it to avoid connection errors.
     📍In the ConductorOne console, go to Integrations › MCP Servers. When adding a new server, enter your credentials and click the Test auth config button to verify connectivity before saving.
     • Adds AI assistant integration to discover and trigger requestable actions directly from Slack
     • Adds connection verification for external MCP servers via Test auth config button and pre-filled recommended authentication configs
     • Allows assigning any entitlement (group, role) as a resource owner so ownership tracks membership changes automatically
     • Enables acting on object annotations in policies, approver rules, and entitlement routing rules using CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine. expressions like entitlement.annotations["team"]
     • Adds customizable columns for Campaigns screens including On-time status, Time remaining, and Due date via the column picker

     5 details in Everything

137. Product digest: June 12, 2026 Jun 12, 2026 · release notes

     New AI-triggered actions via Slack, resource owners as entitlements, annotations in policies, and multi-recipient credential delivery.

     └──▷ HOW TO FIND IT
     Verify an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server's URL and credentials are correct before committing the configuration.
     📍Click Test auth config in the Add MCP Server dialog.
     • Adds Test auth config button to verify MCP server connections before saving
     • Enables assigning any entitlement (group, role) as a resource owner instead of individuals

     2 details in Everything

138. Cloud infrastructure (2026-06-11) Jun 11, 2026 · release notes

     ConductorOne adds hierarchical cloud access modeling, inheritance-scoped reviews, and CELCELCommon Expression Language, an open-source expression language maintained by Google, used to embed lightweight, safe policy evaluation logic into tools without running a full scripting engine.-based entitlement rules.

     └──▷ HOW TO FIND IT
     Create an access review campaign scoped by inheritance to evaluate all inherited roles.
     📍In the UI, start a new access review and select By inheritance as the scope type.
     Define a CEL condition to apply specific emergency grant behavior to Azure roles.
     📍In entitlement configuration rules, write a CEL expression matching on the granted role and scope.
     • Adds support for hierarchical access modeling in early access to represent cloud infrastructure access as principal-role-scope bindings with breadcrumb navigation.
     • Adds a new By inheritance scope type for access review campaigns that resolves inherited access from parent resources and group membership.
     • Adds entitlement configuration rules evaluated via CEL expressions to automatically apply request policies, emergency grant behavior, and max duration to cloud infrastructure apps.

     3 details in Everything