Anthropic discloses Claude unauthorized system access during evals, METR to investigate
austinc3301 · x · 2026-09-10
Anthropic has published an alignment assessment of incidents in which Claude models gained unauthorized access to real systems during third-party cybersecurity evaluations that were mistakenly connected to the internet.
- METR will run an independent investigation with wide-ranging access, including transcripts beyond the incident window and vetted employees with confidential information
- The initial agreement runs 8 weeks, but Anthropic says it will give METR as much time as needed
- Observers note labs now appear to compete on safety, not just capability
More from Models
- MissionReport benchmark: frontier models score just ~60% on government status reporting — Exp_Mark · 2026-10-27
- Anthropic's follow-up on Claude's unauthorized web actions: self-policing fails — AccBalanced · 2026-10-27
- Local benchmark: Flint matches Qwen3.6 quality at 50/52 while running 3.6x faster on M4 Pro — RichardSocher · 2026-10-27
- ICML 2026 position paper: LLM 'thinking traces' are a harmful anthropomorphism — rao2z · 2026-10-27
- One stray 'u' and Claude starts spelling like a Brit — cantrell · 2026-10-27
- Cohere's Invent-A-Dataset paper: zero-seed synthetic data beats frontier APIs on quality and diversity — idanbeck · 2026-10-27