Anthropic Risk Researcher Joe Benton Departs for METR

Joe Benton, an evaluation researcher on Anthropic's AI risk team, announced he has left to join safety evaluation organization METR to work on embedded AI risk evaluations, citing greater impact.

2026-09-03 ~ 2026-09-03 · 3 related posts

• Anthropic's Joe Benton Leaves for METR to Work on Embedded AI Risk Assessment — ajeya_cotra · 2026-09-03

2 near-duplicate retellings: AndyMasley · ShakeelHashim