The United States Justice Department is scrutinising Nvidia’s planned $12.93 billion purchase of Hugging Face, Reuters reported on October 28, 2026, and CNBC later confirmed the inquiry. Investigators want to know whether owning the largest hub for open artificial intelligence models would give the chipmaker leverage. Nvidia declined to comment on the probe, according to CNBC. The inquiry establishes no wrongdoing, and the department could close its work without taking legal action.
Nvidia announced on September 3, 2026 that it would acquire Hugging Face for $12.93 billion, according to TechCrunch and The Hill. The merger agreement was signed a day earlier, per the company’s filing with the Securities and Exchange Commission. The deal is expected to close in the first half of 2027, pending regulatory approval. Hugging Face hosts more than 3 million models and about 500,000 datasets, TechCrunch reported, with more than 18 million developers and more than 200,000 company customers.
Why Nvidia wants Hugging Face
Hugging Face is not a chip company or a model lab. It runs the platform where researchers and companies publish and test open models and datasets. Developers treat it as the default shelf for open weight releases. That position is why antitrust staff care who owns it. At $150 million in annualised revenue, reported by The Information via TechCrunch, the $12.93 billion price works out to roughly 86 times sales. Buyers rarely pay that multiple for revenue alone. They pay for position, and regulators know it.
The company already took part in Hugging Face’s 2023 funding round, which raised $235 million and was led by Salesforce Ventures, with Google, Amazon, and IBM also investing, per TechCrunch. Full ownership would turn a minority stake into control. The chipmaker also ships its own open models, so a hub it owns could favour builds optimised for its graphics processors. That outcome is a hypothetical, and the company denies the intent.
What the Justice Department is examining
Reuters, as relayed by Investing.com, said enforcers are focused on whether absorbing Hugging Face would let the chipmaker take control of a distribution channel for AI models and software. Bloomberg described the department as making inquiries about the agreement and its effect on competition. CNBC said the company has been asked to provide information about the competitive effects.
Channel Insider framed the concern as the chipmaker using ownership to disadvantage competing chipmakers and AI providers. The questions include whether rival hardware gets equal treatment and how Hugging Face tools would work with rival systems. A second line of inquiry covers sales conduct. Channel Insider reported that investigators are looking at whether Nvidia offered preferential access to AI chips when customers also bought networking equipment and Groq processors. That tied selling theory is separate from the Hugging Face deal itself, and nobody has proven it.
Other Nvidia deals under review
Hugging Face is not the first transaction by the company to draw questions. Investing.com listed the Run:ai and SchedMD acquisitions among the deals under examination, along with the Groq arrangement. Bloomberg reported on September 10 that the Justice Department was already probing the Groq licence deal on antitrust grounds. Outlets disagree on the Groq price tag, put near $20 billion by Bloomberg and CNBC and at $17 billion by heise, which described a nonexclusive technology licence plus hires, without buying the company.
Each deal gives Nvidia a piece of the software layer above its chips. Regulators tend to look hard when a dominant hardware supplier keeps buying the tools customers use to run that hardware.
How Nvidia has answered the concern
Nvidia has answered the concern in public. Per Investing.com, it says Hugging Face will keep open support across rival processors and software ecosystems. Heise reported a similar line, that the platform will stay open to competing providers of AI chips, cloud services, and models. At announcement, chief executive Jensen Huang said Hugging Face would continue to back open source and open weight models, according to reports on the September 3 deal. He also said developers would pick their own models and compute, and that the company’s hardware would not be required to build on or deploy through the platform.
Pledges are not contracts. Regulators in other merger reviews have weighed behavioural promises against structural remedies, and they often prefer the second. The practical question is whether the company would accept binding commitments on neutrality. It has not said so publicly. CNBC named Microsoft’s 2018 purchase of GitHub for $7.5 billion as the closest comparison. That deal closed, and the platform stayed broadly open to rival clouds. Skeptics note that Hugging Face costs far more, and that the buyer sells the scarce hardware the whole AI field wants. Investing.com reported that Nvidia shares dropped 4.22 percent to $204.64 after the Reuters story, a snapshot rather than a closing price.