Updates - METR
METR — named in the OpenAI system card for AI self-improvement evaluations; publishes methodology a third party could rerun
8
Lines added
0
Lines removed
Was read
2026-10-21
c1f14b7569b3a5db…
Changed by
2026-10-23
c5c53fa960ac10d7…
8 lines added
Left column is the version read 2026-10-21, right is 2026-10-23. Unchanged stretches are elided. Both bodies are kept in full and are addressed by the hashes above.
⋯ 36 unchanged lines
3737 Menu
3838
3939 Updates
4040
41+Red-Teaming OpenAI's Internal Agent Monitoring Systems
42+
43+October 22, 2026
44+
45+A METR employee and a Redwood Research employee spent three weeks stress-testing OpenAI's internal misalignment monitoring system, discovering a number of novel vulnerabilities.
46+
47+Read more
48+
4149 Review of the "Misalignment in high-stakes settings" section in Anthropic's Risk Report (August 2026)
4250
4351 October 20, 2026
4452