Yann LeCun's AMI Labs Raises $1.03 Billion in Record European Seed Round to Build World Models That Challenge the LLM Paradigm
Turing Award winner Yann LeCun's Paris-based startup raises the largest seed round in European history to develop world models, an alternative AI architecture that learns from physical reality rather than text prediction.
Overview
AMI Labs, the artificial intelligence startup founded by Turing Award winner Yann LeCun after his departure from Meta, has closed a $1.03 billion seed round at a $3.5 billion pre-money valuation, according to TechCrunch. The round, announced on March 10, 2026, is believed to be the largest seed investment ever raised by a European startup and one of the largest globally.
The Paris-headquartered company is building what it calls world models: AI systems that learn from multimodal sensory data and physical environments rather than relying on next-token text prediction, the technique that powers large language models such as GPT and Claude. LeCun has described the current LLM approach as “a statistical illusion” that is “impressive, yes” but “intelligent, no,” as reported by The Next Web.
What We Know
The Funding
AMI Labs initially sought around 500 million euros but demand exceeded that figure significantly, ultimately closing at 890 million euros (approximately $1.03 billion), according to TechCrunch. The round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions.
Strategic investors include chip manufacturer Nvidia, automaker Toyota, Samsung, and Singapore sovereign wealth fund Temasek, as reported by Bloomberg. Individual backers include web inventor Tim Berners-Lee, venture capitalist Jim Breyer, Dallas Mavericks owner Mark Cuban, and former Google CEO Eric Schmidt, according to The Next Web.
The Team
LeCun serves as executive chairman, while Alexandre LeBrun, previously CEO of medical AI startup Nabla, leads the company as CEO. The leadership team includes Laurent Solly, formerly Meta’s vice president for Europe, as chief operating officer; Saining Xie, a former Google DeepMind researcher, as chief science officer; Pascale Fung, formerly a senior director at Meta AI, as chief research and innovation officer; and Michael Rabbat, a former Meta research director, as vice president of world models, according to TechCrunch.
LeCun left Meta in November 2025 after 12 years building the company’s AI research operation, as reported by CNBC.
The Technical Vision
At the core of AMI’s approach is the Joint Embedding Predictive Architecture (JEPA), which LeCun first proposed in a 2022 research paper. Unlike large language models that predict the next word in a sequence, JEPA learns abstract representations of reality from multimodal data, including video, audio, and sensor inputs. The architecture is designed to mimic how humans and animals build internal models of physical reality through embodied experience rather than language alone, according to The Next Web.
The company’s initial product focus centers on an “AMI Video” model, with target markets spanning robotics, manufacturing, and wearable technology, per TechCrunch.
European Positioning
LeCun has deliberately headquartered AMI Labs in Paris, with additional offices planned in New York, Montreal, and Singapore. The company frames itself as a rare frontier AI lab that is “neither Chinese nor American,” according to The Next Web. The decision positions Europe as a potential leader in the emerging world models paradigm before U.S. companies establish dominance, as they have in the LLM space.
AMI Labs has committed to publishing research papers and releasing significant portions of its code as open source, per TechCrunch.
What We Don’t Know
- Whether world models will prove to be a viable commercial alternative to large language models remains unproven. AMI Labs has no product, no revenue, and no near-term commercialization plans, as acknowledged in The Next Web reporting.
- The timeline for demonstrating world models at a level that competes with or surpasses LLMs on practical tasks is uncertain. AMI’s own leadership has described a three-to-five-year horizon for deploying “fairly universal intelligent systems.”
- How AMI’s partnership with Nabla, LeBrun’s former company, will materialize in healthcare AI remains to be seen. Nabla reportedly serves approximately 85,000 clinicians across more than 130 U.S. health systems.
- CEO Alexandre LeBrun has cautioned that “world models” could become an overhyped buzzword, predicting that “in six months, every company will call itself a world model to raise funding,” per TechCrunch. Whether AMI can deliver substance behind the term will be a defining test.
Analysis
The size of the round reflects both LeCun’s personal stature in the field and growing investor appetite for alternatives to the dominant LLM paradigm. LeCun’s 2018 Turing Award, his years leading Meta’s foundational AI research, and his persistent public skepticism of autoregressive language models have given AMI a contrarian thesis that clearly resonated with capital markets.
The strategic investor roster is notable. Nvidia’s participation hedges its bets beyond the LLM training hardware that has driven its recent growth. Toyota and Samsung signal potential industrial applications in automotive and electronics manufacturing where understanding the physical world matters more than generating text.
Still, LeCun is making a long-duration scientific bet in a market that has rewarded fast product iteration. Companies like OpenAI and Anthropic have built multi-billion-dollar businesses around LLMs while world models remain largely theoretical. The first year of AMI’s existence will be dedicated entirely to research, with corporate partnerships only beginning discussions in year two. Whether a billion-dollar seed round can sustain that patient timeline, and whether investors will accept it, will be one of the most closely watched experiments in the AI industry.