Yann LeCun Reportedly Preparing to Leave Meta to Launch His Own Startup
Meta could soon lose one of its most respected artificial intelligence leaders. Yann LeCun, the company’s Chief Scientist and one of the most prominent figures in the field, is reportedly preparing to leave and start his own venture, according to sources cited by the Financial Times.
LeCun, who is also a professor at New York University and recipient of the A.M. Turing Award, is expected to depart in the coming months. Reports indicate he is already in discussions to secure funding for a new startup dedicated to advancing his research on “world models.”
A world model is a form of intelligent system designed to build an internal understanding of its surroundings, allowing it to simulate cause-and-effect situations and anticipate potential outcomes. Other major research groups, such as Google DeepMind and World Labs, are also investing in this technology.
Meta’s AI Shake-Up and the Future of LeCun’s Research
LeCun’s exit would come at a significant moment for Meta, which is currently reshaping its artificial intelligence strategy amid growing competition from OpenAI, Google, and Anthropic.
The company is said to be reorganizing its internal structure after recently hiring more than 50 engineers and researchers from rival organizations to form a new division known as Meta Superintelligence Labs (MSL). In June, Meta also invested $14.3 billion in data-labeling firm Scale AI and appointed its CEO, Alexandr Wang, to lead the new unit.
These structural changes have reportedly created internal friction within Meta’s artificial intelligence group. Some of the newly hired specialists have expressed dissatisfaction with the complexity and bureaucracy of a large corporation, while members of the company’s earlier generative research team have seen their responsibilities reduced.
LeCun’s long-term research under Meta’s Fundamental AI Research Lab (FAIR) has also been affected. FAIR focuses on foundational research that could influence future systems five to ten years down the line. However, recent leadership shifts by Mark Zuckerberg—including efforts to accelerate product-ready model development—have overshadowed this work, especially after the company’s previous model family, Llama 4, fell behind competitors.
Throughout his career, LeCun has been known for his cautious views on how artificial intelligence should be developed and presented. He has often criticized the notion that current systems are close to achieving human-like reasoning or control.
“It seems to me that before ‘urgently figuring out how to control systems much smarter than us,’ we first need to design one smarter than a house cat,” he recently wrote on social media.
Meta did not provide an immediate comment on the matter outside regular business hours.
Also Read:
AI4Bharat Rolls out ‘Indic LLM Arena’ to Benchmark AI Models for Indian Languages
