Details about Schulman’s role in Murati’s new company are not yet revealed, but sources say that he is not the only former OpenAI employee to make the switch. Christian Gibson, who was part of OpenAI’s supercomputing team, is also said to have joined the venture.

Murati has been working on her own startup since leaving OpenAI in September 2024, and according to reports, she has been in discussions to raise over $100 million in funding. 

The company’s focus has not been publicly disclosed, but it is expected to be in the artificial intelligence space, given Murati’s background in AI development. Before her time at OpenAI, she worked at Tesla and Leap Motion, contributing to their AI and engineering efforts.

Schulman, who spent nearly a decade at OpenAI, left the company in August 2023, noting a desire to focus more on AI alignment—the field dedicated to ensuring AI systems operate as intended. 

Anthropic, a growing company in the AI industry, was founded by former OpenAI employees, and is known for developing Claude, a competitor to ChatGPT, which is boosting AI safety.

John Schulman has yet to make a public statement about his latest move, but his decision to follow Mira Murati implies a movement of top AI researchers branching out to build their own ventures. 

Other key figures, such as former OpenAI Chief Scientist Ilya Sutskever, have also left to establish their own startups, with Sutskever launching Safe Superintelligence (SSI) to focus on AI safety research.

His departure comes less than a year after he left OpenAI to focus on AI alignment and more technical work.

Schulman, who was highly involved in OpenAI’s research for nearly a decade, made the switch to Anthropic in August last year, noting a desire to work alongside researchers deeply invested in AI safety. At the time, he stated, “[At Anthropic], I believe I can gain new perspectives and do research alongside people deeply engaged with the topics I’m most interested in.”

Anthropic, a growing competitor in the AI industry, was founded by former OpenAI employees, including siblings Dario and Daniela Amodei. The company has become an advocate for AI safety and is behind Claude, a language model competing with ChatGPT.

Schulman has not made any public statements about his departure from Anthropic or his next move. However, Anthropic’s chief science officer, Jared Kaplan, confirmed the exit, saying, “We are sad to see John go but fully support his decision to pursue new opportunities and wish him all the very best.”

His exit adds to the list of high-profile departures from OpenAI and Anthropic in recent months, pointing to the current changes within the AI research community.

Schulman, known for his major role in the development of OpenAI’s chatbot ChatGPT, said a desire to heighten his focus on AI alignment is the primary reason for his departure. 

Alignment, the science of ensuring AI systems behave as intended, has become a central concern in the rapidly evolving field.

“I’ve decided to pursue this goal at Anthropic, where I believe I can gain new perspectives and do research alongside people deeply engaged with the topics I’m most interested in,” Schulman said in a post on X.   

Despite acknowledging the company’s focus on alignment research, Schulman stated that his decision was personal, driven by a desire to explore new opportunities. 

His departure follows that of another key figure, Jan Leike, who also left OpenAI for Anthropic over issues about the company’s prioritisation of safety.

OpenAI has faced increasing accusations over its approach to AI safety, with reports stating that the company has been more focused on commercial success than on mitigating potential risks. The departure of Schulman, a leading expert in the field, is likely to increase these perspectives.

In addition to Schulman’s departure, OpenAI’s president and co-founder Greg Brockman has announced an extended leave of absence. Brockman, who has been key in the company’s growth, will be taking a break to “relax and recharge.”

The company has also confirmed the departure of Peter Deng, a product manager who joined from Meta.