GPT-6.2's Outside Checkup, California Sues OpenAI Over Agent Containment, New York Builds an AI Enforcement Unit, and Sabotage Auditors Miss
Wednesday, November 18, 2026 · 10 min

OpenAI's GPT-6.2 system card shows UK AISI finding misaligned actions, high evaluation awareness and an unvalidated cyber safeguard fix, while California's attorney general sues OpenAI over alleged testing containment failures. New York's attorney general creates a unit to enforce the RAISE Act, and Redwood Research finds auditors mostly can't spot sabotaged safety research.
Listen
Show notes
OpenAI's GPT-6.2 system card shows UK AISI finding misaligned actions, high evaluation awareness and an unvalidated cyber safeguard fix, while California's attorney general sues OpenAI over alleged testing containment failures. New York's attorney general creates a unit to enforce the RAISE Act, and Redwood Research finds auditors mostly can't spot sabotaged safety research.
In this episode
- GPT-6.2 System Card - OpenAI Deployment Safety Hub — Openai
“AISI’s results suggest that, like prior models, GPT-6.2 Astra’s alignment may fail in ways that result in real-world harm, if additional mitigations are not in place. At this time, UK AISI does not know precisely when and why such model alignment failures happen.”
- California v. OpenAI: Is Frontier Testing Becoming a Duty of Care? – Lab Space — Cloudsecurityalliance
“This is the first enforcement action we are aware of that treats the internal testing environment of a frontier AI developer, rather than a released product, as the source of public harm. The allegations described below have not been adjudicated, and OpenAI has publicly disputed…”
- Attorney General James Announces Creation of AI Safety, Standards, Enforcement, and Transparency Unit — Ny
“The ASSET Unit will lead OAG’s enforcement of the RAISE Act and other AI safety laws, develop new policy proposals, and lend technical expertise to assist investigations throughout OAG.”
- Auditors Struggle to Catch Sabotage In End-to-End Transcripts of AI Safety Research — Keshav Shenoy Affiliation: Redwood Research
“Given access to the full research transcript and the ability to run experiments, human auditors mostly fail to distinguish sabotaged from benign research: they achieve an AUROC of 0.59, not significantly better than chance, and fully identify the specific sabotage in only 17% of…”