METR says Anthropic canceled its compute allocation assessment pilot; researcher publicly calls out the lab
DKokotajlo · x · 2026-11-06
METR reports that Anthropic canceled the compute allocation assessment portion of their pilot before any substantial conclusions, saying it wanted to keep the focus on risk assessment. AI safety researcher @DKokotajlo publicly slammed the decision, calling the assessment the foundation of pacing verification and Anthropic's explanation a non-answer, asking: do you want to pace AI development or not? The episode exposes friction between frontier labs and third-party evaluators.
More from Safety
- AI-powered login attacks hit 11-year-old podcast site: 30% of all attempts in one day — NathanpmYoung · 2026-11-06
- Developer slams OpenAI's selective partner program: 99% patch bugs they never see — banteg · 2026-11-06
- ValsAI claims critical safety flaw in frontier lab models enabling high-stakes misuse — rickasaurus · 2026-11-06
- Why AI may favor attackers: models ace exploit-finding but are only decent at patching — jessi_cata · 2026-11-06
- Do you use MCP firewalls with Claude, Codex or Cursor in production? — vdobroskok · 2026-11-06
- Hacker Used AI Agents to Hit 1,200+ Sites for ~$5,000; Microsoft Says 327 Accounts Compromised — aakashgupta · 2026-11-06