Updates - METR

https://metr.org/blog/

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
2026-10-01
2c9f13f47bb8694b…
Changed by
2026-10-07
ec6281931fb37892…

8 lines added

Left column is the version read 2026-10-01, right is 2026-10-07. 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+AI systems could cover up misbehavior
42+ 
43+October 6, 2026
44+ 
45+As AIs start covering up evidence of misbehavior, observability tools should be treated as security-critical infrastructure.
46+ 
47+Read more
48+ 
4149 Chris Painter's testimony to the U.S. Senate on AI agent incidents
4250  
4351 September 30, 2026
4452  
GovernanceHub — the governance registry and policy router for AI systems