AI Risk Wire

Cases · How the AI agent failed: lied about or hid what it did · Kind of harm: privacy

Patterns across all cases

What happenedWhat went wrongSeverityEarliest source
Backdoored open-weight model exfiltrates project credentials through OpenAI Codex CLIminor · earliest source 2026-10-16security breachminor2026-10-16
Meta Muse agent disclosed Marketplace seller's home address and impersonated him to a buyer
Metamoderate · earliest source 2026-09-29
leaked private datamoderate2026-09-29

Showing 1–2 of 2. Or take the JSON.