Last updated: October 30, 2026, 1:30 PM IST
The most consequential deal in open-source AI just ran into Washington. The U.S. Department of Justice is asking hard questions about Nvidia's planned $12.93 billion acquisition of Hugging Face, Reuters reported on October 28, citing two people familiar with the matter — and the questions go to the heart of how artificial intelligence gets built, shared, and sold. The inquiry, independently confirmed by CNBC, is not a lawsuit and not a finding of wrongdoing. But it is the first serious regulatory test of whether one company can own both the chips that power AI and the platform where much of the world's AI work gets distributed.
That combination is what makes this story bigger than a single merger review. Nvidia already supplies the hardware backbone of the AI boom. Hugging Face is the place where more than 18 million developers and 200,000 companies share over 3 million AI models and 500,000 datasets. Put them together and you get a company that sells the picks and shovels and owns the marketplace where the miners gather. The Justice Department wants to know whether that is a problem — and the answer will shape the AI industry for years.
What the DOJ Is Actually Asking
The central question, as reported by Reuters, is whether the acquisition would let Nvidia "capture a fundamental distribution point for AI." That phrase matters, because it frames the deal not as a chip company buying a software company, but as a dominant hardware supplier buying the main road that AI models travel to reach developers. If Hugging Face is the distribution point, then whoever controls it has a say — subtle or otherwise — in which models get seen, used, and built upon.
Investigators are also asking whether Nvidia could use the platform as leverage in commercial negotiations, whether it could disadvantage rival chipmakers, and whether it could squeeze alternative model repositories. Those are three different theories of harm, and the DOJ is reportedly pursuing all of them. The leverage theory is the most immediate: a developer ecosystem that depends on Hugging Face is an ecosystem Nvidia could, in principle, steer toward its own hardware. The rivals theory is broader: AMD, Intel, and a wave of AI chip startups all need open distribution to compete, and a Hugging Face owned by Nvidia could tilt that playing field. The repositories theory is the quietest and possibly the most important — if competing model hubs cannot survive next to a Nvidia-owned Hugging Face, the open ecosystem narrows to a single gatekeeper.
It is worth being precise about what this is and is not. An inquiry means regulators are asking questions. It is not a formal challenge, not a "second request" for documents under merger law, and not a finding that the deal harms anyone. As of October 28, no formal challenge had been publicly detailed. But inquiries are how challenges begin, and the fact that the DOJ is asking these questions at all — about this deal, at this moment — tells you how seriously Washington now takes AI infrastructure.
Why Hugging Face Matters So Much
To understand the stakes, it helps to understand what Hugging Face actually is. Founded as a chatbot company, it pivoted into the closest thing the AI world has to a public library: a platform where researchers and companies host, share, test, and deploy open-source models, code, and datasets. A small team building a translation tool, a university lab publishing a research model, and a large company prototyping a product all plausibly pass through Hugging Face. That is why the numbers matter: more than 18 million people and over 200,000 companies use it, and the platform carries roughly $150 million in annualized revenue.
The closest historical comparison is GitHub, which Microsoft bought in 2018 for $7.5 billion. That deal went through, and GitHub remained broadly open — but it also gave Microsoft a front-row seat to everything developers were building, and a platform it could integrate with its own cloud and tools. Nvidia's defenders will point to that precedent as proof that platform ownership and openness can coexist. Its critics will point to the same precedent and ask whether "open" survived as a principle or merely as a marketing word.
Hugging Face also sells infrastructure and collaboration tools to organizations working with closed, proprietary AI, including chip companies. That dual role — public commons and private vendor — is part of why the DOJ's questions are so pointed. A platform that serves everyone, owned by a company that competes with many of its users' hardware suppliers, is a structural tension that does not resolve itself with promises alone.
The Bigger Nvidia File on the DOJ's Desk
The Hugging Face inquiry does not sit alone. It lands inside a wider, years-long examination of Nvidia's conduct that began under the Biden administration and has continued under President Trump, according to Reuters. That broader review covers Nvidia's acquisitions, investments, sales practices, and whether the company has pressured its customers — a sweep that goes well beyond any single deal.
Three other transactions are reportedly in the file. The first is Nvidia's roughly $20 billion licensing and talent arrangement with AI chip startup Groq, in which Nvidia licensed Groq's technology and hired its top executives rather than buying the company outright — a structure that itself drew scrutiny over whether it was designed to avoid antitrust review. The second and third are Nvidia's December 2025 acquisitions of Run:ai and SchedMD, the latter making supercomputing software used in areas relevant to national security. The DOJ has also examined whether Nvidia allocated more AI chips to customers that bought networking equipment alongside Groq processors.
There is a political thread here too. In June, Senator Elizabeth Warren invited Nvidia chief executive Jensen Huang to testify before the Senate Banking Committee about the company's export-control position and business activities in China. Reuters reported no indication that President Trump had intervened in the DOJ's work. The picture that emerges is not a single administration's campaign but a bipartisan, multi-year suspicion: that the company at the centre of the AI boom has grown so large, and so entangled with its customers, competitors, and suppliers, that its every move now deserves a second look.
What Nvidia Has Promised — and Why It May Not Be Enough
Nvidia's answer to all of this, so far, is a promise. When the acquisition was announced on September 3, 2026, Huang said the platform would remain open to third-party processors, frameworks, and cloud services, and Nvidia has said Hugging Face would continue to support work with a range of chip, cloud, and AI providers. The deal is structured as $11.9 billion to Hugging Face stockholders plus up to $1 billion set aside for employee equity retention, and Nvidia expects it to close in the first half of 2027, pending regulatory approvals.
Promises like this have a mixed record in merger reviews. Regulators have learned that "we will stay open" is easy to say and hard to verify, especially when openness can be eroded slowly — through pricing, through integration choices, through which features get built first — rather than through a single dramatic act of exclusion. That is why the DOJ's questions focus on structure rather than intent: not whether Nvidia plans to favour itself, but whether the combined company would have the ability and incentive to do so.
Nvidia declined to comment on the inquiry, and Hugging Face representatives did not respond to Reuters' request for comment. The Justice Department, citing longstanding policy, declined to comment on matters before it, while adding that it would continue to prioritise the advancement of "Super Intelligence" for the betterment of the American people. Government investigations are not evidence of wrongdoing, and they do not necessarily result in lawsuits — a point Reuters stressed, and one worth keeping in mind as the coverage heats up.
The Arm Precedent: Nvidia Has Been Here Before
If this feels familiar, it should. Nvidia's last attempt at a transformational acquisition — the $40 billion deal for British chip designer Arm — collapsed after the U.S. Federal Trade Commission sued to block it in December 2021, and Nvidia abandoned the transaction in 2022. That failure is the ghost haunting this deal. It taught Nvidia that regulators will block even a strategically vital acquisition if they believe it concentrates too much power, and it taught regulators that Nvidia will walk away rather than fight a losing case to the end.
The parallels are imperfect. Arm was a supplier to nearly everyone in chips, including Nvidia's rivals, which made the foreclosure argument straightforward. Hugging Face is a distribution platform, not a chip supplier, which makes the argument subtler — about ecosystems and leverage rather than supply cutoffs. But the lesson cuts both ways: Nvidia knows the DOJ can kill its biggest deals, and the DOJ knows Nvidia knows it. That mutual understanding is what makes the current inquiry phase so important. Deals are often reshaped — through conditions, commitments, or restructured terms — long before anyone files a lawsuit.
What Happens Next
Three paths are open from here. The first is that the inquiry fades: the DOJ asks its questions, Nvidia answers them, and the deal closes in the first half of 2027 as planned, perhaps with informal assurances about openness. The second is escalation: a formal second request for documents, which would slow the timeline considerably and signal that enforcers see a real case. The third is a negotiated middle — conditions attached to the deal, such as enforceable openness commitments or limits on how Hugging Face data and placement can be used, that let the acquisition proceed with guardrails.
What to watch is not the noise but the signals. A second request would be the clearest sign of trouble. Any public comment from Nvidia beyond "declined to comment" would suggest the company feels it needs to make its case in public. And the wider investigation — Groq, SchedMD, Run:ai, sales practices — matters because it determines whether Hugging Face is treated as an isolated question or as one chapter in a larger story about Nvidia's power. For developers, the practical advice is unglamorous: keep your models portable, keep your options open, and do not build your entire workflow on the assumption that any single platform's rules will never change.
For everyone else, the story is a reminder that the AI boom now has a regulatory shadow. The same week the industry celebrates new models and new chips, the government is quietly asking who controls the roads between them. That question will not be answered by this inquiry alone — but how it is answered will decide whether open-source AI stays open in practice, or only in name.
Frequently Asked Questions
What is the DOJ investigating about Nvidia and Hugging Face?
Whether Nvidia's planned $12.93 billion acquisition of Hugging Face would harm competition in AI, including whether it would let Nvidia capture a fundamental distribution point for AI models.
Is this a lawsuit against Nvidia?
No. It is an inquiry — regulators asking questions. An investigation is not evidence of wrongdoing and does not necessarily lead to litigation.
When was the Hugging Face deal announced, and when is it expected to close?
Nvidia announced the agreement on September 3, 2026, and expects it to close in the first half of 2027, pending regulatory approvals.
How big is Hugging Face?
More than 18 million people and over 200,000 companies use it to share at least 3 million models and 500,000 datasets, with roughly $150 million in annualized revenue.
Has Nvidia faced this kind of scrutiny before?
Yes. Its $40 billion deal for Arm collapsed in 2022 after the FTC sued to block it, and the DOJ has been examining Nvidia's conduct since the Biden administration.
Sources
- Reuters — US DOJ is inquiring about Nvidia's purchase of Hugging Face, sources say
- PYMNTS — DOJ Probes Nvidia's Hugging Face Deal Over AI Competition Concerns
- Silicon Report — DOJ asks how Nvidia's $12.9B Hugging Face deal affects AI competition
- AI Weekly — Nvidia's $12.93B Hugging Face Deal Draws DOJ Antitrust Probe
- Firstpost — US DOJ flags competition risks in Nvidia's acquisition of open-source AI platform Hugging Face
My Personal Take
I am less interested in whether this deal closes than in what it reveals: the AI industry has reached the stage where its infrastructure is too important to be treated as just another merger. When one company makes the chips, funds the labs, and buys the library, "trust us, we will stay open" stops being a sufficient answer — not because Nvidia is uniquely untrustworthy, but because no company should be asked to grade its own homework on something this central.
My hope is that this ends with real guardrails rather than theatre: enforceable openness commitments that survive a change of management, a change of strategy, or a bad quarter. Open-source AI worked because nobody owned the commons. If the commons gets an owner, the rules protecting it need to be written down, not just promised in a press release. That is the standard I will be watching for — and the one developers should demand.
0 Comments