AMD announced on 28 September 2026 a definitive agreement to acquire World Labs, the AI model and research lab led by Fei-Fei Li, in an all-stock transaction valued at approximately $8.2 billion. AMD says the deal is expected to close by the end of 2026, subject to regulatory approvals and other customary closing conditions.
AMD’s 8-K, filed the same day, dates the merger agreement to 26 September 2026. We missed the deal when it was announced; this is our read of the filing, AMD’s release and the two posts World Labs and Li published that day.
AMD’s release describes World Labs, headquartered in San Francisco, as a developer of “spatial-intelligence models that generate, reconstruct and simulate interactive 3D environments from text, image and video inputs, as well as technology for robotic learning and simulation.”
In her post, Li writes that she founded World Labs at the beginning of 2024 with Ben Mildenhall and Justin Johnson. She says the company recently released Atlas, which she calls “a first of its kind omni model architecture” that “can predict the next view from an input of 2D images”, and that its acquisition of SceniX is building toward robotics simulation.
After closing, AMD says, the World Labs team will continue to focus on AI model research. Li will join AMD as executive vice president and chief scientist, reporting to Lisa Su, and World Labs says Johnson and Mildenhall will work with her to continue leading the team.
What AMD agreed to pay for World Labs: approximately $8.2 billion, all in AMD common stock. The merger agreement was signed on 26 September 2026 and announced on 28 September; AMD expects to close by the end of 2026, and the number of shares is set from AMD’s average price before closing.
Business Pill · PHYSICAL AI
A short explainer of physical AI: models that perceive, simulate and act in the physical world rather than only in text. It teaches the general idea only and says nothing about any company in this story.
The key insight: As we read it, AMD is buying a research team and a view of future workloads more than a product: its release describes the benefit as insight into how workloads are evolving, and neither the release nor the 8-K gives World Labs’ revenue. The price is paid in AMD stock, so World Labs’ holders take AMD shares rather than cash.
How AMD Is Paying
The 8-K says AMD agreed to acquire all of the equity interests in World Labs Technologies, Inc. “for a total purchase price of approximately $8.2 billion to be paid in shares of the Company’s common stock”, subject to customary adjustments.
It also says the number of shares “is not known” yet and will be calculated from the daily volume-weighted average price of AMD stock over the ten consecutive trading days ending two trading days before closing.
AMD says it intends to issue the shares under the private-placement exemptions of Section 4(a)(2) and Regulation D, and that the issuance is not connected with a public offering.
Why Each Side Says It Did the Deal
Su said in AMD’s release: “Building the compute platforms for the next generation of AI requires a deep understanding of how models are evolving.” The release says World Labs’ model expertise will give AMD deeper insight into how workloads are evolving and help shape its future technology roadmaps.
Li’s post gives the other side. “To do this requires scaling our efforts, widening our reach, and getting closer to the hardware,” she writes, adding: “Without having a focused hardware effort, AI is hobbled in efficiency.”
Both texts use the word open. AMD says the deal advances its strategy to deliver AI infrastructure for an open ecosystem, and World Labs says the combined company is committed to an “end-to-end open AI ecosystem spanning hardware, software, platforms, and widely accessible open models.”
A Relationship That Predates the Deal
Li writes that the two companies were already close. “AMD’s CEO, Dr. Lisa Su, was an early investor and believer in our mission,” her post says, and World Labs says the companies began a deep technical partnership last year, starting with model training and inference optimization on AMD GPUs.
Neither AMD’s release nor its 8-K mentions Su’s investment, and none of the texts we read gives its size.
AMD’s AI moves of 2026 as its own releases describe them: the Anthropic partnership for up to 2 GW of AMD Instinct MI450 GPUs (22 July), the agreement to acquire Taalas (6 August) and the agreement to acquire World Labs (28 September, highlighted).
AMD’s Other AI Moves This Year
World Labs is not AMD’s first AI deal of 2026. On 6 August, AMD announced a definitive agreement to acquire Taalas, a Toronto-based designer of specialized AI inference silicon; that release gives no price.
On 22 July, AMD and Anthropic announced a partnership to deploy up to 2 gigawatts of AMD Instinct MI450 Series GPUs in AMD Helios rack-scale systems, with the first gigawatt beginning in the first half of 2027, AMD says.
Taken together, by our reading of AMD’s own releases, in three months the company has expanded its work with Anthropic and agreed to acquire an inference-chip company and a model-research lab. The July release says Anthropic already used AMD Instinct MI355X GPUs, and neither acquisition had closed when the releases were issued.
The Structural Read
AMD’s release frames the deal from the chip side: World Labs’ expertise in developing advanced models will give AMD “deeper insight into how workloads are evolving” and help shape its future technology roadmaps.
Li’s post frames it from the model side. She writes that without a focused hardware effort, AI work like World Labs’ “remains trapped in the digital world.” Her lab’s models generate and simulate 3D environments for uses that include robotics, AMD’s release says.
As we read it, the two framings meet in the same place: a chipmaker that wants to know what the next workloads look like, and a model lab that says it needs to sit closer to the hardware. AMD’s release calls the target areas reasoning, robotics, simulation and physical AI.
Fei-Fei Li, co-founder and CEO of World Labs, in her post of 28 September 2026
“Without having a focused hardware effort, AI is hobbled in efficiency.”
Three Implications
AMD’S CHIP ROADMAP AMD says World Labs’ model expertise will help shape its future technology roadmaps, with Li reporting to Lisa Su as chief scientist after closing.
ROBOTICS AND SIMULATION DEVELOPERS AMD’s release says World Labs builds technology for robotic learning and simulation, and Li writes that its SceniX acquisition is building toward robotics simulation.
USERS OF OPEN MODELS World Labs says the combined company is committed to widely accessible open models; AMD’s release does not say whether those models will run on hardware other than AMD’s.
The framework puts it this way: “The winning strategy in every major platform shift has followed a consistent pattern: Close the SCARCE layer (keep proprietary) + Open the ABUNDANT layer (commoditize)”.
As we read it, AMD and World Labs both describe the models and the ecosystem around them as the open part, while AMD’s release describes its own business as AI-optimized CPUs, GPUs, networking and software.
What Is Not Established
The texts we read do not say how many AMD shares will be issued, what World Labs’ revenue or last private valuation was, which regulators must approve the deal, or how large Su’s early investment was and how it is being handled. AMD’s release does not say whether World Labs’ models will stay available to developers who do not use AMD hardware.
AMD has agreed to pay about $8.2 billion in its own stock for a model lab that, by World Labs’ account, began a technical partnership with AMD last year on model training and inference optimization on AMD GPUs.
What it says it gets is a view of where AI workloads are heading and a chief scientist, Li, reporting to its chief executive; what the filing leaves open is the share count, which is set only at closing.
A note on sourcing. On 12 October 2026 we read in full AMD’s Form 8-K filed on 28 September 2026, AMD’s press release of the same day, World Labs’ blog post and Fei-Fei Li’s post on her Substack, and AMD’s releases of 22 July and 6 August 2026 on Anthropic and Taalas, all linked below. Every figure, date and quotation comes from those texts; the reading of AMD’s year is ours.
We missed the deal when it was announced on 28 September and are covering it two weeks later. Nothing here is investment advice.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur, HackerNoon and Search Engine People. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
