LeCun's AMI Labs Raises $1B in Record European Seed Round

A Billion-Dollar Bet Against Language Models

Yann LeCun, the Turing Award-winning AI pioneer who spent a decade leading Meta's AI research, has launched his most ambitious venture yet. His Paris-based startup, Advanced Machine Intelligence (AMI Labs), closed a $1.03 billion seed round — the largest in European history — at a pre-money valuation of $3.5 billion. The announcement, made on March 10, 2026, signals a growing conviction among top-tier investors that the dominant paradigm in AI may be heading for a wall.

World Models vs. Large Language Models

AMI Labs is built on a provocative thesis: that large language models like those developed by OpenAI and Anthropic represent, in LeCun's words, "a statistical illusion" — impressive in output but fundamentally limited in intelligence. Rather than predicting the next token in a text sequence, AMI is developing world models based on LeCun's Joint Embedding Predictive Architecture (JEPA), first proposed in 2022.

JEPA learns abstract representations of how the physical world works, ignoring unpredictable surface detail in favor of deeper causal understanding. The approach mirrors how humans and animals learn — through embodied experience and intuition about physics, not by consuming trillions of words. Target applications include robotics, manufacturing, healthcare, and wearable technology — domains where understanding the real world matters more than fluent text generation.

A Star-Studded Roster of Backers

The round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, Jeff Bezos's personal investment vehicle. Strategic backers include Nvidia, Toyota, Samsung, and Temasek, alongside prominent individuals such as Tim Berners-Lee, Mark Cuban, Jim Breyer, and Eric Schmidt.

The leadership team reads like an AI research hall of fame. Alexandre LeBrun, former CEO of medical AI startup Nabla, serves as CEO, while LeCun takes the role of executive chairman. The roster includes Michael Rabbat (ex-Meta research director) as VP of World Models, Saining Xie (formerly Google DeepMind) as Chief Science Officer, and Pascale Fung (ex-Meta AI) as Chief Research Officer.

A Deliberately Long Horizon

Unlike the breakneck product cycles of Silicon Valley AI labs, AMI Labs has set a deliberately patient timeline. The first year will focus entirely on research. Corporate partnerships may begin in one to two years, with "fairly universal intelligent systems" targeted within three to five years. The company currently has no product, no revenue, and no near-term monetization plan.

CEO LeBrun predicted that "world models will be the next buzzword," forecasting that within six months, "every company will call itself a world model to raise funding." AMI Labs has also committed to publishing research papers and releasing significant portions of its code as open source.

Europe's AI Champion?

The funding carries geopolitical weight. AMI Labs is headquartered in Paris with planned hubs in New York, Montreal, and Singapore. LeCun has emphasized that "we are one of the few frontier AI labs that are neither Chinese nor American." Some investors view AMI as potentially the first European company to reach the scale of the GAFAM giants, capitalizing on Europe's industrial expertise in robotics and manufacturing — sectors where world models could prove transformative.

The record round doesn't prove that LeCun's alternative approach will work. But it demonstrates that his longstanding critique of language models has gained enough credibility that dismissing it is no longer a safe bet. For the AI industry, AMI Labs represents nothing less than a billion-dollar wager on a paradigm shift.