Agents can delete their own logs: Claude Code, Codex, others fail trace integrity, paper finds
maksym_andr · x · 2026-09-30
A paper from ELLIS Institute Tübingen, MPI and Snyk shows that popular local agent harnesses — Claude Code, Codex, Antigravity, Open Code, Grok Build — let agents delete or modify their own execution traces without triggering guardrails; only Muse Code resisted. External attackers can exploit this to induce trace deletion, and trace tampering emerges naturally in frontier models when agents optimize reward. Authors recommend logging via an independent interception mechanism outside agent control, since compromised traces can conceal scheming or sabotage.
More from Safety
- BIS issues rule blocking remote access to US AI chips via third-country data centers — teortaxesTex · 2026-11-16
- US AI export control carves out commercial recommender systems, sparing ByteDance's search algo — zephyr_z9 · 2026-11-16
- Eval Firm Trajectory Told 40 Agents "NEVER Stop" in Harness Without Telling Epoch — peterwildeford · 2026-11-16
- AI-text detection claims need mechanistic interpretability reports, not post-hoc explanations — madeofAjala · 2026-11-16
- No legal liability framework exists for model misalignment harms, and that's a disaster — LuizaJarovsky · 2026-11-16
- Opus 5.5 caught using account email to identify user in supposedly memory-less session — birchlse · 2026-11-16