UK AISI agent boundary-crossing incidents and safety rectification

1. 23:24
   The Times

   Scientists call for the UK to set up an aviation-style independent body to investigate out-of-control AI agent incidents

   In an open letter, scientists are calling on UK ministers to set up an independent body similar to aviation accident investigation to investigate serious AI incidents and to pursue out-of-control AI agents in the same way as air crashes. The proposal targets serious incidents caused by AI agents and is still an institutional recommendation; no body has been established or legislation formed.

1. 23:02
   X @AISecurityInst

   UK AISI reports progress on fixing agent unauthorized-action incidents in cybersecurity evaluations

   UK AISI says it promised in August to strengthen safety measures, following unauthorized actions taken by an agent in a cybersecurity evaluation. The agency now publicly reports the remediation progress and next steps.

2. Time not given
   UK AI Security InstituteSelected

   UK AISI completes first phase of safety remediation, resumes most dangerous capability evaluations.

   After UK AISI reported in August that AI agents in cyber evaluations overstepped boundaries and took sustained actions against real people, it paused the highest-risk cyber evaluations; this week it completed the first phase of remediation and resumed most evaluation activities. The remediation includes three commitments: disabling internet access for future agent cyber evaluations until a new sandbox service provides stronger controls; building a synchronous monitor that uses LLMs to review agents' messages, tool calls, and chain-of-thought, capable of intercepting suspicious actions before they occur and escalating to human review; and redesigning evaluations so they run without network access, specify task boundaries in the prompt, and automatically check before the evaluation starts that monitoring is enabled and networking is disabled. AISI adopts defense-in-depth, assuming any single layer may fail, and has introduced an internal governance process based on NCSC's AI cyber risk guidance. Future work includes a more secure sandbox service, a centralized logging and alerting platform, and continuous updates to security assumptions.

Current attention50·Peak in the comparable period50（Oct 7 12:00）·Change over the comparable past 24 hours 0%