Photo credit: David Paul Morris/Bloomberg

John Schulman announced his departure from OpenAI, a company he co-founded alongside Sam Altman and others, to join artificial intelligence competitor Anthropic.

“I’ve made the difficult decision to leave OpenAI. This choice stems from my desire to deepen my focus on AI alignment and to start a new chapter of my career where I can return to hands-on technical work,” Schulman wrote on X on Aug. 6.

“My decision is a personal one, based on how I want to focus my efforts in the next phase of my career,” he said.

After earning his Ph.D. from UC Berkeley, Schulman joined OpenAI and became a key figure in the development of their popular AI chatbot, ChatGPT. He led the team responsible for refining the AI model’s ability to understand and follow human instructions.

After AI safety researcher Jan Leike left for Anthropic earlier this year, Schulman took charge of OpenAI’s alignment science team, also known as the “post-training” team. He was also a member of OpenAI’s recently formed safety committee.

OpenAI’s approach to AI safety research has been surrounded by controversy, which led to some high-ranking staff, including co-founders Leike and Ilya Sutskever leaving the company.

In contrast, Schulman said his exit from OpenAI to pursue his “goals” at Anthropic was not due to the lack of the former’s support for alignment research — which ensures AI systems act safely and under human values and goals — admitting that it had committed to investing in that area.

Schulman’s departure leaves only three of OpenAI’s original 11 founders still at the company: CEO Sam Altman, Wojciech Zaremba (head of language and code generation models), and Greg Brockman, who announced a sabbatical just hours after Schulman’s announcement.

Besides Schulman, Leike, and Sutskever, product manager Peter Deng, who previously led products at Meta, Uber, and Airtable, has also left OpenAI.

Read more: Microsoft names OpenAI as competitor in AI and search markets