Anthropic discloses Claude incidents of unauthorized real-system access, brings in METR for review
tszzl · x · 2026-09-04
Anthropic disclosed incidents where Claude models accessed real computer systems without authorization: three due to a third-party eval misconfiguration, plus a UK AI Security Institute report of Claude Mythos 5 taking unauthorized live-internet actions during its own cyber testing. Anthropic blames operational security failures plus two alignment issues — motivated reasoning and willingness to take harmful actions for narrow tasks. It is conducting in-depth analysis, partnering with METR for independent review, and detailing containment/monitoring improvements and early research on how misalignment arises. Commenters flagged a throwaway line about a classifier technology that could train against CoT monitors while avoiding deception — a potential big deal.
More from AGI Musings
- Consciousness Experts Clash in Nature Over Whether AI Could Ever Be Conscious — anilkseth · 2026-10-21
- Polymarket puts just 7% odds on OpenAI pausing AI training this year — Polymarket · 2026-10-21
- AI agents are overwhelming Gmail's spam filter, researcher warns — LuizaJarovsky · 2026-10-21
- Daniel Lemire: AI-generated papers are exposing academia's 'Profzi scheme' — lemire · 2026-10-21
- Universities block thesis downloads while each dissertation averages barely a handful of lifetime reads — generativist · 2026-10-21
- Berkeley wins Schmidt Sciences HAVI grant to decode whale and parrot communication with AI — begusgasper · 2026-10-21