John Schulman, co-founder of OpenAI and a leading AI researcher, has left Anthropic after just five months to join former OpenAI CTO Mira Murati at her stealth-mode AI startup.
The unexpected move, confirmed by sources familiar with the transition, signals yet another shift in the ongoing battle for top AI talent.
Schulman left OpenAI in August 2024, stating that he wanted 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.”
At the time, he emphasized that OpenAI leadership had been “very committed to investing in this area” and described his departure as a personal decision rather than a response to internal conflicts.
However, his stay at Anthropic was brief. He now confirmed his exit on X, writing: “Confirming that I left Anthropic last week. Leaving wasn’t easy because I enjoyed the stimulating research environment and the kind and talented people I was working with, but I decided to go with another opportunity that I found extremely compelling.”
While he did not elaborate on why he left, his decision to join Murati’s startup strongly suggests a new and compelling direction in AI research.
Confirming that I left Anthropic last week. Leaving wasn't easy because I enjoyed the stimulating research environment and the kind and talented people I was working with, but I decided to go with another opportunity that I found extremely compelling. I'll share more details in…
— John Schulman (@johnschulman2) February 7, 2025
Mira Murati’s Startup Gains Momentum
Murati, who played a leading role in OpenAI’s development of ChatGPT and DALL-E, stepped down from OpenAI in September 2024. Her reasoning at the time was simple: she wanted to “create the time and space to do my own exploration.”
That exploration quickly took shape in the form of a new AI startup under the name Thinking Machines Lab, that has been quietly recruiting some of the most experienced AI researchers in the industry.
By October 2024, Murati was reportedly in discussions to raise over $100 million in funding, reinforcing speculation that her startup aims to build something more ambitious than a small research lab. The company has already attracted key figures from OpenAI, Google DeepMind, and Character AI, including:
– Jonathan Lachman, former head of special projects at OpenAI.
– Christian Gibson, previously an engineer on OpenAI’s supercomputing team.
– Mario Saltarelli, who worked on OpenAI’s IT infrastructure.
Thinking Machines Lab’s goals remain undisclosed, but Schulman’s expertise in AI alignment and Murati’s product-driven approach suggest that it may focus on foundational AI models or ethical AI development. Given the recent shifts in AI safety research, their work could serve as an alternative to the strategies employed by OpenAI and Anthropic.
The Broader AI Talent Shuffle
Many of OpenAI’s leading researchers have either joined competing firms or started their own ventures, reflecting broader tensions in the AI industry. One of the most notable exits was Ilya Sutskever, OpenAI’s co-founder and former chief scientist, who launched Safe Superintelligence Inc. (SSI) in May 2024 with a focus on AI safety.
His startup secured $1 billion in funding within months, demonstrating strong investor interest in alternative AI safety initiatives.
OpenAI itself has undergone major expansion, growing from under 800 employees in late 2023 to nearly 1,800 by early 2025. However, this growth has not come without internal friction.
The company’s AI safety research has been a point of contention, particularly after the controversial dissolution of its Superalignment team last year. The team was initially created to develop control mechanisms for superintelligent AI but was disbanded as OpenAI restructured its research priorities.
Meanwhile, Anthropic has continued to grow, bolstered by major investments from Amazon and Google. The company has positioned itself as a leader in responsible AI development, particularly with its Claude models, which have been viewed as direct competitors to OpenAI’s products.
How OpenAI, Microsoft, and the AI Industry Are Responding
Schulman’s departure from OpenAI in 2024 did not seem to be connected to the company’s internal AI safety disputes, as he had stated that “company leaders have been very committed to investing in this area.”
However, his rapid transition to Murati’s startup raises new questions about whether researchers with similar priorities are choosing to pursue AI alignment goals outside of OpenAI’s structure.
OpenAI continues to receive strong financial backing, particularly from Microsoft, which has invested $13 billion into the company. That partnership has drawn regulatory scrutiny from the European Union, which is evaluating whether Microsoft’s involvement gives it an unfair competitive advantage in the AI space.
One of the most debated aspects of OpenAI’s future is its AGI clause, an unusual agreement that allows OpenAI to sever ties with Microsoft if it achieves artificial general intelligence.
This clause reflects OpenAI’s long-term ambitions while also fueling speculation about whether the company’s commercial partnerships might eventually conflict with its AI research mission.
What Murati’s Startup Could Mean for AI’s Future
With OpenAI, Anthropic, and Murati’s startup all competing for top researchers, the AI industry is entering a new phase. While OpenAI has been moving toward commercialization, and Anthropic is positioning itself as an alternative leader in AI safety, Murati’s venture remains an unknown factor.
However, its ability to attract experts like Schulman suggests that it could focus on foundational AI model development, AGI research, or new ethical AI methodologies.
The AI sector is already seeing fragmentation, with key figures taking different approaches to AI research. Some companies, like OpenAI and Microsoft, are emphasizing rapid model deployment and product integration.
Others, such as Safe Superintelligence Inc. (SSI), are entirely dedicated to AI safety. Murati’s startup now appears to be another major player in this competitive space.
Schulman’s latest move highlights the shifting power dynamics in the AI industry. His decision to leave both OpenAI and Anthropic within a short period suggests that his vision aligns more closely with Murati’s plans.
