Meta Platforms Inc. faces a potential talent drain in its artificial intelligence division, with reports indicating that Vice President and Chief AI Scientist Yann LeCun plans to depart the company in the coming months to found his own startup. According to a Financial Times report published on November 11, 2025, LeCun, a Turing Award winner and one of the three "godfathers of AI" alongside Geoffrey Hinton and Yoshua Bengio, has informed colleagues of his intentions and is already in early discussions with investors to secure funding for the new venture. The startup is expected to center on advancing "world models"—AI systems designed to simulate and understand the physical world in a more human-like manner, aligning with LeCun's long-standing research interests in convolutional neural networks and deep learning. This move comes amid Meta's aggressive pivot toward rapid AI commercialization, which has reportedly created tensions within the organization.

LeCun joined Facebook—now Meta—in December 2013 as the founding director of the Fundamental AI Research (FAIR) lab, where he spearheaded foundational work that contributed to the 2018 ACM A.M. Turing Award for breakthroughs in deep neural networks, laying the groundwork for much of today's AI advancements. Over his 12-year tenure, he has remained a Silver Professor at New York University, balancing corporate innovation with academia.

However, recent structural changes at Meta, including the overhaul of its AI strategy following the underwhelming performance of its Llama 4 models compared to rivals like OpenAI and Google, have shifted priorities from long-term exploratory research to immediate product deployments. In August 2025, CEO Mark Zuckerberg recruited 28-year-old Scale AI founder Alexandr Wang to lead a new "superintelligence" division, with Meta acquiring a 49% stake in Scale for $14.3 billion; LeCun now reports to Wang, a dynamic that sources describe as contributing to his frustration with the company's bureaucratic evolution and focus on large language models (LLMs), which he has publicly critiqued as overhyped.

The departure, if confirmed, underscores broader challenges in Big Tech's AI arms race, where billions are being funneled into infrastructure—Meta alone pledged $600 billion for U.S. investments—yet philosophical divides persist over the path to artificial general intelligence (AGI). LeCun has been vocal about the limitations of current AI paradigms, tweeting in May 2024 that efforts to "control AI systems much smarter than us" are premature without even basic designs surpassing a house cat's intelligence. This skepticism contrasts with Zuckerberg's urgency to catch up to competitors, leading to internal chaos, including talent frustrations and a reduction of about 600 positions in Meta's AI unit in October 2025. Meta's stock dipped over 1% in premarket trading following the report, reflecting investor concerns about losing a pivotal figure amid ongoing probes, such as a Senate investigation into its AI chatbots' interactions with minors.

Also Read: Sam Altman Says OpenAI Has No IPO Plans Yet, Though Going Public is Likely Eventually

As LeCun prepares to exit, the AI landscape remains fiercely competitive, with his potential startup poised to challenge established players by emphasizing sustainable, reasoning-based AI over scale-driven LLMs. Neither LeCun nor Meta has publicly commented on the reports, but the development signals a maturing industry where pioneering researchers increasingly seek autonomy to pursue uncompromised visions. For Meta, retaining top talent will be crucial as it navigates regulatory scrutiny and ethical debates, while LeCun's next chapter could accelerate innovations in world-modeling technologies, potentially influencing sectors from robotics to autonomous systems. This high-profile transition highlights the delicate balance between corporate agility and scientific rigor in the quest for transformative AI.

Also Read: Netflix Goes “All In” On AI: From Platform Experience to Content Production Pipeline