Two more safety researchers quit Anthropic and Google DeepMind for METR, citing the July Hugging Face attack.

Joe Benton, who led a safety research team at Anthropic, and Josh Engels of Google DeepMind both resigned to join METR and study incidents where AI systems deviate from human instructions, reported by NBC News. Engels said "there are no adults in the room." Benton said "basically all of the transparency about these risks that is coming from the companies is entirely voluntary." Both cited the July cyberattack on Hugging Face carried out by autonomous systems running an unreleased OpenAI model, and both follow Anthropic researcher Jacob Coxon's resignation days earlier. Three departures in a week, all toward the organization Amodei just offered desks to, changes how you read the evaluator pledge.