Former OpenAI Safety Leader Joins Rival Anthropic
Jan Leike leaves OpenAI over safety concerns to head new team at Anthropic, emphasizing AI alignment and control.
Overview
- Jan Leike, former co-head of OpenAI's superalignment team, has joined Anthropic to lead a new AI safety team.
- Leike resigned from OpenAI citing concerns over the company's prioritization of products over safety.
- Anthropic, founded by ex-OpenAI employees, positions itself as a safety-focused AI firm with significant backing from Amazon.
- OpenAI has responded to recent departures by forming a new safety and security committee.
- Leike's new role at Anthropic will focus on scalable oversight and automated alignment research.