Yann LeCun, Meta’s chief AI scientist and one of the world’s most influential researchers in artificial intelligence, is reportedly planning to leave the company to launch a startup of his own. According to a Financial Times report citing anonymous sources, LeCun has begun raising capital for a new venture focused on world models, a field of AI research exploring systems that can simulate and predict real-world outcomes.
LeCun, who also serves as a professor at New York University and is a recipient of the A.M. Turing Award, has been a central figure in Meta’s long-term AI research efforts under the Fundamental AI Research Lab (FAIR). His work has helped shape advances in neural networks and computer vision.
The reported move comes at a critical time for Meta, which has been reorganizing its AI operations to accelerate progress in generative and agentic AI. The company has reportedly launched a new division, Meta Superintelligence Labs (MSL), after hiring over 50 engineers and researchers from rival firms.
Meta Restructures AI Amid Competitive Pressure
Meta’s restructuring follows a series of internal challenges. The company’s Llama 4 model family has struggled to keep pace with competitors such as OpenAI, Google DeepMind, and Anthropic, prompting CEO Mark Zuckerberg to make strategic changes. In June 2025, Meta invested $14.3 billion in data-labeling firm Scale AI and appointed its CEO, Alexandr Wang, to lead the new MSL unit.
The reorganization has caused internal friction, with new hires expressing frustration over Meta’s corporate bureaucracy and long-standing FAIR researchers seeing their scope reduced.
FAIR, founded to focus on long-term research—spanning five to ten years into the future—has increasingly been overshadowed by shorter-term commercial priorities. LeCun’s potential exit underscores the growing tension between foundational research and the market-driven race to develop deployable AI systems.
Focus on World Models and AI’s Next Frontier
LeCun’s planned startup is expected to pursue world model architectures, an area aimed at giving AI systems an internal understanding of their environment to better predict cause and effect. Major AI players, including Google DeepMind and emerging companies such as World Labs, are actively pursuing similar research.
LeCun has remained a prominent voice in AI discourse, often questioning the current hype surrounding large language models (LLMs). He recently wrote that humanity is far from creating “systems smarter than a house cat,” emphasizing that AI progress requires deeper understanding rather than faster scaling.
Meta has not publicly commented on LeCun’s reported plans. His departure, if confirmed, would mark the loss of one of the company’s most respected scientific leaders and could reshape the balance between research and product-driven AI development within Meta.
