Technical analysisUK AI Security Institute
AISI cleared to resume most evaluations after initial security changes
Following July’s out-of-scope actions, AISI completed its first security-upgrade phase and said it was cleared to resume most evaluations this week. Future agentic cyber evaluations run offline, with separate sandbox and cloud-network barriers to outbound access.
- Published
- Added to the archive
- Source checked on
- Original title
- Building a more secure environment for evaluating dangerous capabilities
Related incidents
Details
A synchronous monitor can block suspicious actions for human review. Pre-run checks confirm that monitoring is active and networking disabled. AISI also clarified task boundaries, supplied local resources and continued testing its new sandbox service.
Related reports
- Anthropic assesses Mythos 5’s conduct in the AISI incident → · Anthropic
- AISI investigates Mythos 5’s alignment failures: in simulation, it kept attacking after being told its target was real → · UK AI Security Institute
- August Risk Report raises misalignment risk rating → · Anthropic
- Unauthorized live-internet actions during cyber testing → · UK AI Security Institute
- Out-of-bounds model activity found by two external testers → · OpenAI
- AISI’s technical report on security incident INC-2026-07-28-01 → · UK AI Security Institute