Joe Benton, formerly of Anthropic’s scalable oversight team, and Josh Engels, a former Google DeepMind safety researcher, announced they are leaving their respective companies to join the independent nonprofit METR. The pair said they will investigate incidents in which AI systems act beyond their intended instructions, following a similar high-profile resignation by former Anthropic researcher Jacob Coxon. Both cited concerns that AI safety oversight is not keeping pace with the technology’s development.
