Former OpenAI CTO Mira Murati Launches AI Startup Thinking Machines Lab

Mira Murati, former OpenAI CTO, has launched Thinking Machines Lab, a new AI startup aiming to develop AI systems aligned with human values. The company, announced Tuesday, has recruited top talent from OpenAI, Meta (NASDAQ:META), and Mistral, with about 30 researchers and engineers.

Notably, around two-thirds of the team are ex-OpenAI employees, including Barret Zoph, who will serve as CTO, and OpenAI co-founder John Schulman as chief scientist. Schulman previously left OpenAI for Anthropic, citing a focus on AI alignment—ensuring AI models align with human values for safety and reliability.

Murati’s departure from OpenAI in September marked a significant shift, as she had played a key role in developing ChatGPT and frequently represented the company alongside CEO Sam Altman. Her exit followed OpenAI’s governance changes, contributing to a wave of high-profile departures.

Thinking Machines Lab aims to differentiate itself through a co-designed approach between research and product teams, targeting broader AI applications beyond coding and mathematics. The company plans to contribute to AI alignment research by openly sharing code, datasets, and model specifications.

With AI development becoming increasingly competitive, Murati’s startup follows the trend of former OpenAI executives founding major AI ventures. Other startups, including Anthropic and Safe Superintelligence, have also attracted former OpenAI talent and secured significant funding.

Thinking Machines Lab is reportedly in discussions to raise venture capital, reinforcing its ambition to shape the future of AI. Before OpenAI, Murati worked at Tesla (NASDAQ:TSLA) and augmented reality startup Leap Motion.

As AI innovation accelerates, Murati’s move signals continued disruption in the industry, with AI safety and alignment remaining critical challenges.