Meta’s chief AI scientist, Yann LeCun, is reportedly planning to leave to build his startup

Meta may be about to lose one of its most famous AI heads: Yan Liqun, the company’s chief AI scientist, is planning to leave the company to build his startup, the Financial Times reported, citing anonymous sources.

The report added that LeCun, a professor at New York University, senior researcher at Meta, and winner of the prestigious AM Turing Award, is planning to leave in the coming months, and is already in talks to raise capital for a startup that will focus on continuing his work on global models.

A global model is an artificial intelligence system that develops an internal understanding of its environment so that it can simulate cause-and-effect scenarios to predict outcomes. Top labs and startups like Google DeepMind and World Labs are also developing global models.

LeCun’s departure will come at a pivotal time for Meta, which recently changed how it approaches AI development in response to concerns that it is being outpaced by competitors such as OpenAI, Google and Anthropic.

The company has reportedly begun revamping its AI organization after hiring more than 50 engineers and researchers from its competitors to build a new AI unit, dubbed Meta Superintelligence Labs (MSL). It is worth noting that in June, Meta invested $14.3 billion in data classification company Scale AI, and appointed its CEO Alexander Wang to manage the new division.

These decisions have made things increasingly chaotic at Meta’s AI unit, as new talent expressed frustration at dealing with the bureaucracy of a large company, while Meta’s previous generative AI team saw limited scope, sources told TechCrunch in August.

CEO Mark Zuckerberg’s decisions to fix things have overshadowed LeCun’s long-term research work at the company within its Fundamental AI Research Laboratory (FAIR) division after the company’s previous family of AI models, Llama 4, failed to keep up with competing models. Unlike MSL, FAIR is designed to focus on long-term AI research, technologies that could be used five to ten years from now.

LeCun has been openly skeptical about how AI technology — specifically LLMs — is currently being marketed as a cure-all for humanity’s ills. He even tweeted that AI systems have a long way to go.

“It seems to me that before we urgently figure out how to control AI systems smarter than ourselves, we need to get a head start on designing a system smarter than a house cat,” he wrote.

Meta did not immediately respond to a request for comment outside regular business hours.