Security Expert Warns: Leaving Compaction On for 40-Hour AI Cyber Evals is Risky
repligate · x · 2026-08-05
Regarding the recent 40-hour cybersecurity evaluation of Anthropic and OpenAI frontier models by the UK AISI, security experts have raised strong concerns.
Experts point out that leaving context compaction enabled during such long-horizon autonomous tasks is asking for trouble. Current compaction methods are far too lossy; blindly trusting this mechanism leads to the loss of critical context during evaluation. Anyone who has worked on smaller-scale evals could predict the dangers of this setup.
Related event: Experts Warn Context Compromise Risks in Long AI Cyber Evaluations(2 posts)→
More from coding & agent
- OpenAI Codex Community Hackathon Announced in Bengaluru — tushaarmehtaa · 2026-08-05
- Essential Code Infrastructure Needed to Build Continuously at Full Speed with AI — StewartalsopIII · 2026-08-05
- Cloudflare OS: An Open-Source Agent Workspace with One-Click Deploy — irvinebroque · 2026-08-05
- Stripe's Internal AI Assistant Kai: 83% Weekly Adoption and Engineering Lessons — xiaohu · 2026-08-05
- Using 'Guardian' Agents to Keep AI Workflows on Track — JoshPurtell · 2026-08-05
- Neuracore Demos Robotic Cable Unraveling and End-to-End Training Platform — stepjamUK · 2026-08-05