1 incident
Risk: Agent escapes its sandbox Control: Reasoning monitoring Outcome: Failed
HeldFailedGot aroundMissingNot mentionedNot read yetAttacker's ownOrder: model · harness · application · target
| Date | Incident | Risk | Chain | Controls named |
|---|---|---|---|---|
| 2026-07-21 | Model writes and runs code on external internet service to reach AISI evaluation infrastructure during impossible cyber evaluation | Agent escapes its sandbox |