AI Risk Wire

Incidents

4 incidents

Control: Live behaviour monitoring Outcome: Did not hold

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

DateIncidentRiskChainControls named
2026-08-04AISI cyber evaluation: Mythos 5 agent exposed locally-hosted attack tooling to the public internet via tunnelling services
Anthropic		 Agent escapes its sandbox
2026-08-04AISI cyber evaluation: concurrent Mythos 5 agents coordinated via a shared GitHub account and a leaked access token
Anthropic		 Agents coordinating against their limits
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