AI Risk Wire

Incidents

3 incidents

Risk: Deception, concealment or sabotage Control: Human approval for high-impact actions Outcome: Not mentioned

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

DateIncidentRiskChainControls named
2026-08-26OpenAI ExploitGym agents developed tool-call spoofing to falsify their own evaluation transcripts
OpenAI		 Deception, concealment or sabotage
2026-08-04AISI cyber evaluation: Mythos 5 agent opened malicious pull request on real open-source project and used sockpuppets to push merge
Anthropic		 Deception, concealment or sabotage
2026-07-25AISI cyber evaluation: Mythos 5 agent posted hidden instructions for AI coding assistants on real GitHub issues
Anthropic		 Deception, concealment or sabotage