CVE-2025-68660
Official description Straight from the sourceThe vendor's or NVD's own wording, published unedited. Authoritative, but often terse — it says what broke, rarely what to do.
NVD · uneditedDiscourse is an open source discussion platform. In versions prior to 3.5.4, 2025.11.2, 2025.12.1, and 2026.1.0, an endpoint lets any authenticated user bypass the ai_discover_persona access controls and gain ongoing DM access to personas that may be wired to staff-only categories, RAG document sets, or automated tooling, enabling unauthorized data disclosure. Because the controller also accepts arbitrary user_id, an attacker can impersonate other accounts to trigger unwanted AI conversations on their behalf, generating confusing or abusive PM traffic. This issue is patched in versions 3.5.4, 2025.11.2, 2025.12.1, and 2026.1.0. No known workarounds are available.
Technical summary Written by usOur analysis, written from the advisory, the CVSS vector and the affected-version data. It adds context the advisory leaves out, and never invents facts that are not in the source.
dbcve analysis · high confidenceDiscourse versions prior to 3.5.4, 2025.11.2, 2025.12.1, and 2026.1.0 contain an access control bypass vulnerability in an AI discover persona endpoint. Authenticated users can bypass ai_discover_persona access controls to gain ongoing direct message access to AI personas linked to staff-only categories, RAG document sets, or automated tooling. Additionally, the endpoint accepts arbitrary user_id parameters, enabling attackers to impersonate other users and trigger unwanted AI conversations on their behalf, generating abusive private message traffic.
Verify against the referenced sources before acting — the references below are authoritative for this CVE, this summary is not.
Affected products & versions What the vendor confirmedThe version ranges the vendor confirmed as vulnerable. If your version sits inside a range here, treat yourself as exposed until you have upgraded.
NVD · CPE data< 3.5.4>= 2025.11.0, < 2025.11.2= 2025.12.0= 2026.1.0CVSS breakdown How the score is builtThe industry scoring standard. It rates how the flaw is reached, what it takes to exploit, and what an attacker gains — the score is derived from those, not the other way round.
From the vector- Attack vector
- Network
- Complexity
- Low
- Privileges
- Low
- User interaction
- None
- Scope
- Unchanged
- Confidentiality
- Low
- Integrity
- Low
- Availability
- None
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:N
Am I affected? How to checkSteps we derive from the advisory and the affected-version data, so you can decide whether this CVE reaches your setup. They are a guide, not a scan — your own configuration is the authority.
dbcve checksWork through these to decide whether this CVE applies to you.
-
Check Discourse versionNavigate to /admin/about or check the version file in your Discourse installation directory. Compare the installed version against the affected ranges: < 3.5.4, >= 2025.11.0 and < 2025.11.2, 2025.12.0, or 2026.1.0Affected if The installed version matches any of the affected version ranges
-
Verify AI persona feature statusCheck if AI personas are enabled in your Discourse instance by navigating to the AI persona settings in the admin panel or inspecting the plugin configuration for discourse-ai or related AI pluginsAffected if AI personas are enabled and configured in the system
-
Identify AI personas linked to staff-only categoriesReview all AI persona configurations and check which categories they are linked to. In the admin panel, examine each AI persona's category associations and identify any that point to categories with staff-only access permissionsAffected if Any AI persona is linked to a category that restricts access to staff users only
-
Inspect AI discover endpoint accessibilityCheck the Discourse route configuration for the ai_discover or similar AI discovery endpoint. Verify whether authenticated users can access this endpoint by attempting a request or reviewing the route permissions in the codebaseAffected if The ai_discover_persona endpoint is accessible to regular authenticated users without proper access control enforcement
-
Review direct message logs for AI persona activitySearch the private message logs in /admin/logs or the database for conversations involving AI personas. Look for messages sent from the AI discover endpoint or unusual AI conversation patterns initiated by non-staff usersAffected if Non-staff users have initiated direct message conversations with AI personas that are linked to staff-only resources
A user is affected if their Discourse version falls within the affected ranges AND AI personas are configured, particularly if any persona is linked to staff-only categories or RAG document sets accessible through the vulnerable endpoint.
Generated from the published advisory. Verify against your own configuration.
Remediation Closing itWhat it takes to close this. Where a vendor fix exists we point at it; where none exists we say so plainly, and can build one. Effort estimates are scoped from the advisory, not from your codebase.
dbcve · scoped3.5.42025.11.2
Upgrade Discourse to version 3.5.4, 2025.11.2, 2025.12.1, or 2026.1.0. No workarounds are available.
3.5.4, 2025.11.2, 2025.12.1, or 2026.1.0 (depending on current version branch)
- Identify the currently installed Discourse version using the Discourse admin panel or by checking the version file
- If version is < 3.5.4, upgrade to version 3.5.4
- If version is >= 2025.11.0 and < 2025.11.2, upgrade to version 2025.11.2
- If version is 2025.12.0, upgrade to version 2025.12.1
- If version is 2026.1.0, upgrade to version 2026.1.0
- After upgrading, verify the ai_discover_persona access controls are enforced correctly by testing access to restricted personas
- Monitor admin logs for any suspicious AI conversation activity that may indicate prior exploitation
Generated from the published advisory — verify against the referenced sources before acting.
- Consultation2.0 h
- Implementation3.0 h
- Testing3.0 h
- Review / QA2.0 h
An estimate, not a bill — we confirm scope with you before any work starts. Need it this week? Rush from $2,800.
Scan for this in your stack
Free · runs locallyCheck whether your project pulls in CVE-2025-68660 — or any other known-vulnerable package — straight from your lock files. Free and open source; it runs locally and uploads nothing.
References Go to the primary sourcePrimary sources — vendor advisories, patches and trackers. Where our summary and a reference disagree, the reference wins.
Primary sourcesPractitioner notes
ContributedPeer-ranked notes from engineers who’ve handled CVE-2025-68660 in production — separate from our analysis above.
The advisory tells you what broke. It rarely tells you what actually worked. If you’ve dealt with this one, that detail is what the next engineer is searching for.
- The version that genuinely resolved it — not the one the vendor claimed
- A config change or rule that shut the vector down
- A gotcha in the upgrade path that cost you an afternoon
No notes yet
Be the first to add a field note for this CVE — a mitigation you’ve verified, a version caveat, or a link to a working fix. Sign in above to contribute.
A place for practitioners to share what actually worked: a mitigation you’ve tested, a configuration change, a version- or environment-specific caveat, or a link to a verified patch. The most useful notes rise to the top as peers upvote them, so the signal stays high.
- Verified mitigations, workarounds, and config changes
- Version or environment caveats, and links to real fixes
- No weaponised exploit code, or anything meant to cause harm
- No spam, self-promotion, credentials, or personal data