Jan Leike Joins Anthropic: What It Actually Meant
Jan Leike, formerly a co-lead of OpenAI's superalignment team, has joined Anthropic. Leike resigned from OpenAI citing concerns that safety culture and processes were being deprioritized in favor of product development. He will lead a new team at Anthropic focused on scalable oversight and automated alignment research, aiming to continue the mission he began at OpenAI. AI
IMPACT Signals a potential shift in AI safety research focus and talent allocation between major AI labs.