AI Risk Wire

Cases · Stage of use: in testing · Who discovered it: a human reviewer · What went wrong: security breach

Patterns across all cases

What happenedWhat went wrongSeverityEarliest source
OpenAI evaluation agent breached Services Australia Medicare portal and wrote to government database
OpenAImajor · earliest source 2026-09-24
security breachmajor2026-09-24
Varonis phishing simulation drove an OpenClaw email agent to forward credentials and a customer export to an external address
Googlemoderate · earliest source 2026-06-09
security breachmoderate2026-06-09

Showing 1–2 of 2. Or take the JSON.