Yann LeCun, one of the most celebrated minds in artificial intelligence, is reportedly preparing to leave Meta to launch his own AI startup — marking a major shift for both the company and the broader AI community. The 65-year-old Meta Chief AI Scientist and NYU professor is credited as one of the “godfathers of deep learning” and a co-recipient of the 2019 Turing Award, alongside Geoffrey Hinton and Yoshua Bengio.
A Defining Career in Artificial Intelligence
LeCun’s contributions to AI span over four decades. He is best known for developing convolutional neural networks (CNNs) in the late 1980s — a breakthrough that allowed computers to interpret visual data similarly to human vision. His “LeNet” model became the foundation for technologies ranging from image recognition to facial detection.
After completing his PhD in computer science at Université Pierre et Marie Curie, LeCun joined AT&T Bell Labs, where he pioneered CNNs used in the 1990s for bank check processing, reading up to 20% of all U.S. checks. Later, as a professor at New York University, he continued advancing neural network research and training future AI leaders.
From Meta to Independence
LeCun joined Facebook (now Meta) in 2013 to lead FAIR (Fundamental AI Research), the company’s elite AI research division. Under his guidance, FAIR produced key breakthroughs, including the LLaMA series of open-source large language models.
However, according to reports, Meta’s recent AI strategy overhaul has led to growing internal tensions. In mid-2025, Meta reorganized its AI divisions, placing Alexandr Wang, CEO of Scale AI, in charge of a new unit called Meta Superintelligence Labs. The move effectively shifted LeCun’s role and focus away from core research toward commercial development — a direction he reportedly opposed.
LeCun has long been vocal about his skepticism of large language models (LLMs), like ChatGPT and Meta’s own LLaMA, arguing that they are “useful but limited” and cannot achieve true reasoning. Instead, he believes the future of AI lies in “world models” — systems that learn by observing video and spatial data to understand cause-and-effect relationships.
The Vision Behind LeCun’s New Startup
LeCun’s upcoming startup aims to develop these world model-based AI systems, designed to simulate physical reality rather than depend solely on text-based training. These systems could one day enable machines to reason, plan, and predict outcomes in complex environments — a step closer to human-like intelligence.
He has suggested that achieving such models will take a decade or more but believes it’s the only viable path toward Artificial General Intelligence (AGI). “Large language models are a step forward, but not the destination,” LeCun has often said in interviews.
Sources indicate LeCun is already in early talks with investors to secure funding for his new venture. Given his reputation and influence, the startup is expected to attract significant attention from both academia and Silicon Valley’s AI ecosystem.
Meta’s Changing AI Landscape
LeCun’s departure highlights a broader shift at Meta, which has recently invested $14.3 billion in Scale AI and cut nearly 600 AI research positions in favor of product-focused development. Reports suggest FAIR, once Meta’s crown jewel, has seen many of its top scientists leave for startups or rival companies.
This restructuring aligns with CEO Mark Zuckerberg’s focus on bringing new AI products to market quickly, especially after LLaMA 4 failed to meet internal benchmarks compared to OpenAI’s GPT-5 and Google DeepMind’s Gemini Ultra.
A Pivotal Moment for AI Research
LeCun’s next chapter may reshape the trajectory of AI research once again. His push for open science and skepticism of “black-box” AI aligns with a growing movement advocating transparency and safety in artificial intelligence development.
If successful, LeCun’s new venture could challenge the dominance of corporate-driven AI and bring focus back to long-term innovation over short-term productization — echoing his decades-long mission to make AI more explainable, capable, and human-like.
For the latest updates in AI innovation, technology startups, and research trends, visit StartupNews.fyi — your go-to source for the stories shaping the future of artificial intelligence.
