Yann LeCun, one of the most influential figures in artificial intelligence, is reportedly set to leave Meta to form an independent AI startup.
As reported by TechSpot, citing the Financial Times, the new venture will focus on advancing world models, systems designed to help AI reason about and predict real-world dynamics. LeCun is said to be in early discussions to raise funding for the company.
LeCun has served as Meta’s chief AI scientist since 2013, when he founded Facebook AI Research (FAIR). Under his direction, the lab became one of the most respected centers for foundational AI work, including breakthroughs in self-supervised learning and autonomous systems.
His earlier creation of LeNet in the late 1980s became the prototype for convolutional neural networks that now underpin most modern computer vision systems.
Meta’s shares fell by about 1.5 percent in pre-market trading following reports of LeCun’s departure, underlining investor unease over leadership changes within its research group.
The move also coincides with Meta’s larger structural transition toward product-driven AI development.
Meta’s Expanding AI Portfolio
LeCun’s potential exit comes at a time when Meta is scaling its AI investments aggressively. The company has pledged more than 600 billion dollars in U.S. AI infrastructure and development by 2028.
It also recently invested 14.3 billion dollars to acquire a 49 percent stake in Scale AI, bringing in founder Alexandr Wang to lead parts of its AI effort.
Reports indicate that LeCun now reports to Wang, reflecting a realignment of Meta’s AI hierarchy.
The company has also undergone workforce restructuring. In October, Meta laid off around 600 employees in its AI units, saying the move would streamline decision-making. FAIR and AI infrastructure divisions were affected, though Meta Superintelligence Labs, focused on foundation models, remained untouched.
LeCun has often voiced skepticism about the narrative of AI superintelligence and existential risk, calling such predictions “ridiculous” and “complete B.S.” His more grounded approach contrasts sharply with the escalating hype cycles around AGI and may influence the tone of his next venture.
LeCun’s departure, if confirmed, would signal the end of an era for Meta’s research program and could reorient the global AI landscape. His planned company is expected to explore the next frontier of machine intelligence beyond the text-based focus of large language models, pushing toward systems capable of reasoning, planning, and self-understanding.
For Meta, it raises the question of how to sustain long-term research culture while racing to commercialize AI at scale. For the wider field, it marks a possible return to independent, theory-led exploration at a time when corporate AI research is increasingly bound to product imperatives.
