Artificial intelligence pioneer and Turing Prize laureate Yann LeCun has successfully secured $1.03 billion in seed funding for his new startup, Advanced Machine Intelligence (AMI Labs). Launching shortly after his departure from Meta last year, the Paris-based company is focused on developing AI world models that learn from real-world physical experiences rather than relying solely on language inputs.
This substantial financial injection gives the newly formed enterprise a pre-money valuation of $3.5 billion. It instantly propels the startup to unicorn status before a single product has been shipped to the market. The massive capital raise highlights the growing industry appetite for alternative artificial intelligence architectures capable of understanding complex physical environments.
Moving Beyond Traditional Language Models
The primary goal of AMI Labs is to create systems that can reason and plan within complex real-world settings. Current artificial intelligence approaches largely rely on generative models, such as the widely used Transformer architecture, which focus on predicting the next word in a sentence or the next pixel in an image. LeCun has openly critiqued these limitations, stating that simply predicting text or pixels will not produce broadly capable intelligent agents.
Instead, the new venture will focus entirely on AI world models. These advanced systems are designed to process data from cameras and other physical sensors. By learning abstract representations of real-world sensor data, the technology can effectively ignore unpredictable or irrelevant details. This allows the models to make accurate predictions in a representation space rather than focusing on minute, pixel-level data.
Furthermore, these action-conditioned AI world models will enable agentic systems to predict the real-world consequences of their physical actions. This capability allows the software to plan complex sequences of actions to successfully accomplish specific tasks, all while remaining subject to strict safety guardrails.
Real-World Applications Across Industries
Unlike conversational chatbots, the technology under development at AMI Labs targets highly complex physical industries. The startup envisions customers applying its software to fields such as hardware design, robotics, manufacturing, and healthcare. For instance, the models could analyze aircraft component designs to discover new optimization opportunities.
In the robotics sector, the company is creating systems that can predict the future state of an environment. This technology could empower industrial robots to plan the most efficient way to complete physical tasks, such as properly preparing and moving shipping parcels.
The startup has already secured its first major partnership with Nabla, a clinical artificial intelligence developer. Through this collaboration, Nabla will receive early access to the new technology to develop next-generation agentic tools designed specifically for demanding healthcare workflows. The long-term objective of AMI Labs is to build a general-purpose world model capable of automating difficult tasks across multiple global industries.
A Historic European Seed Round
The $1.03 billion, or roughly €890 million, funding milestone is widely reported as the largest seed round in European history. The French laboratory initially aimed to raise €500 million late last year. However, high investor interest and the prestige of the founding team ultimately pushed the final total significantly higher.
The historic seed round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Jeff Bezos’ Bezos Expeditions. The round also attracted participation from a diverse, globe-spanning roster of corporate backers, including Nvidia, Samsung Electronics, Toyota Ventures, Temasek, Sea, Bpifrance, Groupe Industriel Marcel Dassault, Association Familiale Mulliez, Publicis Groupe, and Aglæ Ventures. Several prominent individual investors also contributed, including World Wide Web inventor Tim Berners-Lee, telecommunications billionaire Xavier Niel, and former Google executive Eric Schmidt.
Celebrating the achievement, LeCun noted on social media that the company had just completed one of the largest seed rounds ever, adding that the startup is actively hiring new talent.
Pioneering Background and Future Plans
The leadership team driving this vision has deep roots in the technology sector. Before founding the new startup, LeCun served as the chief of artificial intelligence at Meta. He originally joined the company in 2013 to establish Facebook AI Research, where he oversaw the development of the ubiquitous PyTorch framework and the Llama language model series. He is also renowned for his early 1988 work on LeNet, an image processing model family that incorporated the backpropagation algorithm still used in modern neural networks today.
During his tenure at Meta, LeCun developed an architecture known as JEPA, which specialized in ignoring irrelevant input details and storing complex, higher-level data. This foundational concept heavily influences the architectural direction of his new venture. AMI Labs is also building a strong executive team, bringing on former Meta Vice President for Europe Laurent Solly to serve as the chief operating officer.
Moving forward, the company plans to release its first models quickly. The leadership team also intends to open-source select pieces of its technology and publish academic papers, ensuring that the broader research community benefits from its innovations.
