Executive brief
OpenClaw is an AI automation platform that integrates with Google Chat for collaboration and task execution. A flaw in its authorization logic allowed attackers to gain unauthorized access to protected resources by manipulating space display names. This could permit unauthorized users to access confidential information or perform actions they should not be permitted to do.
Technical details
The vulnerability is a classic authorization bypass (CWE-639, CWE-807, CWE-863) stemming from reliance on mutable input (space display names) to enforce access control decisions. Google Chat group policies were keyed on displayName rather than immutable group identifiers, allowing an attacker to either change a display name to match an authorized group or create a collision where multiple groups share the same display name. This requires authentication (PR:L) and network access (AV:N) but can be exploited without user interaction. The fix, committed in version 2026.3.25, requires stable group IDs for all authorization decisions instead of relying on user-mutable display names. Patch availability: fixed in OpenClaw 2026.3.25 and later.
Affected products
- OpenClaw openclaw <= 2026.3.24
Timeline
- 2026-03-29: disclosed
- 2026-03-25: patched: First patched version 2026.3.25
References
- https://github.com/openclaw/openclaw/security/advisories/GHSA-52q4-3xjc-6778
- https://github.com/openclaw/openclaw/commit/11ea1f67863d88b6cbcb229dd368a45e07094bff
- https://github.com/openclaw/openclaw
- https://www.vulncheck.com/advisories/openclaw-authorization-bypass-via-group-policy-rebinding-with-mutable-space-displayname