AI startup AMI Labs secures $1.03 billion to develop next-generation world models that help machines understand real-world environments.
A new artificial intelligence startup founded by renowned AI researcher Yann LeCun has secured massive funding to pursue a bold vision for the future of artificial intelligence. The company, AMI Labs, recently raised $1.03 billion in funding at a $3.5 billion pre-money valuation, positioning itself among the most heavily funded AI research startups in the world.
The French-based AI lab aims to develop a new generation of artificial intelligence known as “world models.” Unlike traditional AI systems that rely mainly on language data, world models attempt to understand and learn from the physical world and real-world experiences. The funding round highlights growing investor interest in AI technologies that move beyond current large language models and into deeper forms of machine intelligence.
What Are AI World Models?
The concept of world models is closely linked to the research of Yann LeCun, who is widely regarded as one of the pioneers of modern artificial intelligence. As a Turing Award winner and former chief AI scientist at Meta, LeCun has long argued that current AI systems are limited because they mainly rely on text-based training data.
Most popular AI systems today are based on large language models (LLMs), which learn patterns from massive datasets of written information. While these systems can generate convincing text, they often lack a deeper understanding of the real world.
World models aim to solve this problem by allowing AI to learn from real-world environments, sensory data, and physical interactions rather than just written language. According to AMI Labs CEO Alexandre LeBrun, this shift could fundamentally change the capabilities of artificial intelligence. He told TechCrunch that world models could soon become the next major trend in AI innovation.
“My prediction is that ‘world models’ will be the next buzzword,” LeBrun said. “In six months, every company will call itself a world model to raise funding.”
Despite the humorous remark, LeBrun believes AMI Labs has a genuine technological advantage because its research is focused on deep scientific development rather than short-term product launches.
Building AI That Understands Reality
The goal of AMI Labs is ambitious. The company wants to create AI systems capable of understanding how the real world works. To achieve this, the lab plans to develop AI systems based on JEPA (Joint Embedding Predictive Architecture), a concept proposed by Yann LeCun in 2022.
JEPA focuses on allowing machines to predict and understand relationships between events and objects, rather than simply generating text responses. This approach could allow AI to perform complex tasks such as:
• Understanding physical environments
• Predicting real-world outcomes
• Learning from visual and sensory data
• Making decisions based on context
Such capabilities could make AI systems far more useful in industries where accuracy and real-world understanding are critical.
Healthcare as the First Major Application
One of the earliest sectors expected to benefit from world models is healthcare. AMI Labs’ first disclosed partner is Nabla, a digital health startup where CEO Alexandre LeBrun currently serves as chairman.
LeBrun previously led Nabla and experienced firsthand the limitations of large language models in medical environments. In healthcare, even small AI errors—often referred to as “hallucinations”—can have serious consequences. Because of this, the healthcare industry requires AI systems that can understand context and real-world conditions more reliably.
World models could eventually help doctors by:
• Analyzing complex patient data
• Predicting disease progression
• Supporting clinical decision making
• Reducing diagnostic errors
However, LeBrun acknowledged that these capabilities will take time to develop.
A Long-Term AI Research Project
Unlike many AI startups that quickly launch products and generate revenue, AMI Labs is focusing on fundamental research first. LeBrun emphasized that building world models is a long-term scientific challenge rather than a short-term business opportunity.
“It’s not your typical applied AI startup that can release a product in three months, have revenue in six months, and make $10 million in annual recurring revenue in 12 months,” he explained. Instead, the research could take years before practical commercial applications emerge. Despite the long timeline, investors appear confident that world models represent the next major frontier in artificial intelligence.
Massive Investor Interest in World Model Startups
AMI Labs is not the only company exploring this new AI category. Interest in world models has increased significantly among venture capital investors. Recently, other startups in the field have raised major funding rounds. For example, SpAItial secured a $13 million seed round, which is unusually large for a European early-stage AI startup.
Meanwhile, AI pioneer Fei-Fei Li launched a company called World Labs that raised $1 billion in funding to develop similar technology. With its $1.03 billion investment, AMI Labs now joins this small but rapidly growing group of AI research companies focused on world models.
A Team of High-Profile AI Experts
One reason AMI Labs attracted such a large investment is its exceptionally strong team of researchers and executives. The leadership team includes several well-known figures from the global AI research community. The company’s chairman is Yann LeCun, while the CEO role is held by Alexandre LeBrun.
Other key team members include:
• Laurent Solly – Chief Operating Officer and former vice president for Europe at Meta
• Saining Xie – Chief Science Officer
• Pascale Fung – Chief Research and Innovation Officer
• Michael Rabbat – Vice President of World Models
This combination of experienced researchers and industry leaders helped convince investors that the startup could tackle one of the most complex challenges in artificial intelligence.
Global Research Centers
AMI Labs plans to build its research teams across several major technology hubs around the world. The company will focus on four key locations:
• Paris – company headquarters
• New York City – where LeCun teaches at New York University
• Montreal – an important AI research center
• Singapore – to access talent and connect with Asian markets
LeBrun said the company will prioritize quality over quantity when hiring researchers, focusing on recruiting top AI scientists rather than rapidly expanding staff numbers.
Backed by Major Investors and Tech Giants
AMI Labs’ funding round attracted a wide range of investors, including venture capital firms, tech companies, and prominent technology leaders.
The round was co-led by:
• Cathay Innovation
• Greycroft
• Hiro Capital
• HV Capital
• Bezos Expeditions
The startup also received support from major technology companies, including:
• NVIDIA
• Samsung
• Sea Limited
• Temasek
• Toyota Ventures
Notable individual investors include:
• Mark Cuban
• Eric Schmidt
• Tim Berners-Lee
• Rosemary Berners-Lee
• Jim Breyer
This diverse investor group provides AMI Labs with both financial resources and industry connections.
Commitment to Open Research
Despite the intense competition in artificial intelligence, AMI Labs plans to follow a more open research strategy than many modern AI companies. LeBrun said the company intends to publish research papers regularly and release a significant portion of its code as open source.
This philosophy reflects the values of Meta AI Research (FAIR), where both LeCun and LeBrun previously worked. According to the company, open research can accelerate innovation by allowing researchers around the world to collaborate and build on each other’s work. “We think things move faster when they’re open,” LeBrun explained. “It’s in our best interest to build a community and a research ecosystem around us.”
The Future of AI Beyond Language Models
The rise of world models could represent the next phase of artificial intelligence development. While current AI systems excel at generating text, images, and code, they often struggle to understand how the real world actually works. World models aim to close this gap by giving AI systems the ability to learn from reality itself.
If successful, this technology could transform fields such as robotics, healthcare, scientific research, and autonomous systems. For now, however, the technology remains in its early stages. AMI Labs’ billion-dollar funding round signals that investors believe the future of AI may extend far beyond today’s generative models—and that understanding the world itself could be the next breakthrough.
About The Author
