AMD acquires AI company World Labs from Fei-Fei Li for $8.2 billion
AMD announced Monday that it has reached an agreement to acquire World Labs, a San Francisco-based AI lab developing a kind of model called the World Model, which can be used to simulate 3D environments.
AMD announced Monday that the lab was acquired for approximately $8.2 billion in an all-stock transaction. The chipmaker had already invested in World Labs.
For AMD, this is the chipmaker’s second-largest purchase ever, after the roughly $50 billion paid for Xilinx in 2022.
AMD is making a splash in the AI M&A market at a time when the biggest tech companies are opening their portfolios to top engineering and research talent, even for relatively nascent companies.
OpenAI bought Jony Ive’s io startup for about $6.4 billion last year to tap the famous iPhone designer. Nvidia spent a combined $33 billion on AI chipmaker Groq and open source platform Hugging Face, buying the former’s assets in December and agreeing to acquire the latter this month. And Meta paid $14 billion in 2025 for a minority stake in Scale AI in a deal that saw the startup’s founder, Alexandr Wang, create a new AI division.
AI researchers expect that global models will be able to contribute to the development of robots and other physically grounded artificial intelligence applications. AMD plans to separate World Labs from its chipmaking business until the transaction is finalized later this year.
The lab’s research into next-generation models will allow AMD to plan what its AI chips should be capable of years in advance.
World Labs was founded by AI researcher Fei-Fei Li, who is widely considered an AI pioneer as a professor at Stanford and previously worked for Google, conducting AI research.
Li will become AMD’s chief scientist and executive vice president of the company.
In a demo presented by Li and AMD CEO Lisa Su earlier this year, the two executives showed a World Labs model called Marble creating a three-dimensional scene from a few images.
“Intelligent agents, whether robots, vehicles or even tools, can learn in physics-rich digital worlds before they even need to be deployed in the real world, making them much safer,” Li said during the presentation.
Li, in a social media post Monday, called Su a “great friend.”
“We are steadfast in our mission, and even more excited to continue building a world-leading exploratory research organization that is sustainable for decades to come,” Li wrote.
Gn headline