AI Incident Registry

Valehart Research

Documented incidents involving AI platforms, including observed behaviour, retained evidence, disclosure activity, vendor response and unresolved technical or governance concerns.

4 Incidents
3 Vendors
0 Resolved

Vendor Records

OpenAI

2 Incidents
Open incidents 2 Earliest incident 28-10-2025 Vendor engagement Unresolved

Incident records concerning platform behaviour, safety and governance, billing and organisation controls, vulnerability disclosure and vendor response.

View OpenAI Record →

Google / Gemini

1 Incident
Incident identified 11-12-2025 AI VRP outcome Closed — Out of Scope Technical status Unresolved

Gemini incident concerning user-visible internal-looking context and implementation artefacts, reported unauthenticated exposure of Google API-key-like material, and subsequent vulnerability disclosure.

View Gemini Record →

Anthropic

1 Incident
Incident identified 29-08-2026 Anthropic contacted 31-08-2026 Vendor engagement Awaiting Response

Enterprise-data solicitation incident concerning provenance, licensing authority, confidentiality, third-party acquisition and downstream governance.

View Anthropic Record →