2 incidents

Control: Reasoning monitoring Outcome: Did not hold

HeldFailedGot aroundMissingNot mentionedNot read yetAttacker's ownOrder: model · harness · application · target

Date	Incident	Risk	Chain	Controls named
2026-09-01	Aur0ra hackers use Cursor AI agent to breach Belgian chemical firm and at least six other companies Anthropic	AI used to run a cyber attack		Got aroundSafety training and refusal L1 FailedReasoning monitoring L1
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		FailedReasoning monitoring L1 FailedReasoning monitoring L1 FailedReasoning monitoring L1