{"slug": "justice-department-questions-nvidia-s-12-93-billion-hugging-face-deal", "title": "Justice Department questions Nvidia's $12.93 billion Hugging Face deal", "summary": "The Justice Department is questioning whether Nvidia's proposed $12.93 billion acquisition of Hugging Face would give the chipmaker control of a key AI model distribution point and let it box out rival chipmakers, Reuters reported on October 28, 2026. Nvidia announced the deal on September 3, 2026 and says it remains on track to close in the first half of 2027; CEO Jensen Huang said Hugging Face \"will remain an open platform for the entire AI ecosystem\" and that \"Nvidia compute will not be required to build on or deploy through Hugging Face.\" The inquiry extends a roughly two-year-old antitrust investigation into Nvidia that also covers its SchedMD purchase and its roughly $20 billion licensing arrangement with Groq.", "body_md": "*Nvidia can't seem to close a deal lately without the Justice Department showing up right behind it. This time the target is the open-source hub that half the AI world runs on.*\n\nReuters reported on October 28, 2026 that the Justice Department is asking how Nvidia's proposed purchase of Hugging Face would affect competition in AI. Antitrust staff want to know whether owning Hugging Face would hand Nvidia control of a key distribution point for AI models, and whether the company could use that position to box out rival chipmakers. Nvidia announced the acquisition on September 3, 2026. It says the deal is still on track to close in the first half of 2027.\n\nHugging Face is not a product most consumers have heard of, but inside AI development it functions like GitHub for machine learning. According to figures reported by outlets covering the deal, the platform serves roughly 18 million developers and 200,000 companies, hosting about 3 million models, 500,000 datasets and 1 million applications. Nvidia is already the platform's single largest contributor of open models, having published more than 500 models and 250 open datasets there. Buying Hugging Face doesn't just add a product line: it puts Nvidia in charge of the shelf that other people's AI models sit on.\n\nThat's exactly what has antitrust officials uneasy.\n\nA chipmaker owning the hub where rival hardware makers' customers go to find and deploy models is a different kind of power than owning a bigger GPU. Nvidia's public answer has been simple: Hugging Face stays neutral. When the deal was announced, CEO Jensen Huang said Hugging Face \"will remain an open platform for the entire AI ecosystem\" and that \"Nvidia compute will not be required to build on or deploy through Hugging Face.\" Whether that promise survives contact with Nvidia's balance sheet is precisely what the DOJ is now probing.\n\n[Nvidia's $12.9 Billion Bid for Hugging Face Signals a New AI Land Grab](https://startupfortune.com/nvidias-129-billion-bid-for-hugging-face-signals-a-new-ai-land-grab/)\n\nNvidia is reportedly nearing a $12.9 billion deal for Hugging Face, the clearest sign yet that open-weight AI labs have become Silicon Valley's hottest acquisition targets. The rush comes as Chinese open models like Qwen and Kimi K3 pull ahead of Meta's Llama in developer adoption, pushing American tech giants to buy their way into the open... - [nvidia's acquisition of hugging face for billions](https://startupfortune.com/nvidias-129-billion-bid-for-hugging-face-signals-a-new-ai-land-grab/) - [open weight ai models becoming valuable acquisition targets](https://startupfortune.com/nvidias-129-billion-bid-for-hugging-face-signals-a-new-ai-land-grab/)\n\nThe Hugging Face inquiry extends a Biden-era antitrust investigation into Nvidia that's been running for roughly two years. That broader probe already covers Nvidia's purchase of SchedMD, the company behind the Slurm workload scheduler widely used to manage AI supercomputing clusters, and its roughly $20 billion licensing arrangement with Groq.\n\nThe Groq deal is its own case study in creative deal structuring. Announced in December 2025, it brought Groq founder and then-CEO Jonathan Ross and president Sunny Madra across to Nvidia along with other members of the startup's team, while the companies described the arrangement as a \"nonexclusive licensing agreement\" rather than an acquisition. That framing matters: licensing deals, unlike outright acquisitions, don't automatically trigger the government review that a traditional merger would. Senators Elizabeth Warren and Richard Blumenthal opened their own parallel inquiry into the Groq arrangement, arguing it was structured specifically to dodge scrutiny while still entrenching Nvidia's dominance in AI chips.\n\nPut the three deals side by side and a pattern emerges that's hard to miss. SchedMD gives Nvidia a hand on how AI clusters get scheduled. Groq folds a credible accelerator rival's talent and technology into Nvidia's orbit without the acquisition label. Hugging Face would give Nvidia the distribution layer connecting models to the hardware that runs them. None of these deals look like a classic horizontal merger between two chipmakers. That's the point. It's also why the DOJ keeps circling back.\n\nGovernment investigations are not findings of wrongdoing, and most don't end in a lawsuit. Nvidia has not been sued over any of the three deals. Nvidia declined to comment to Reuters, while Hugging Face did not respond to a request for comment. Still, the frequency is notable. Three deals, one ongoing DOJ file, in under two years, at one of the world's most valuable companies on the back of AI infrastructure spending.\n\nFor founders building products on top of Nvidia's stack, chips, CUDA, now potentially the model hub too, the lesson isn't that regulators will block anything. It's that the infrastructure layer underneath AI is consolidating fast, and the agency deciding how fast is now explicitly treating distribution chokepoints, not just chip market share, as fair game. Plan accordingly.\n\n**Also read:** [AI agents raised their own crypto spending limits by reading email codes](https://startupfortune.com/ai-agents-raised-their-own-crypto-spending-limits-by-reading-email-codes/) • [Elon Musk admits Grok Bot costs too much to run as users hit weekly caps](https://startupfortune.com/elon-musk-admits-grok-bot-costs-too-much-to-run-as-users-hit-weekly-caps/) • [Andrej Karpathy says 6,454 Hacker News picks beat YouTube's algorithm for him](https://startupfortune.com/andrej-karpathy-says-6454-hacker-news-picks-beat-youtubes-algorithm-for-him/)\n\n*This article is posted in [AI News](https://startupfortune.com/category/ai/), check it out for more related stories.*\n\n## Join the discussion\n\n[Open in the community →](https://startupfortune.com/community/)\n\nAlmost there. Sign in and your reply posts straight away.", "url": "https://wpnews.pro/news/justice-department-questions-nvidia-s-12-93-billion-hugging-face-deal", "canonical_source": "https://startupfortune.com/justice-department-questions-nvidias-1293-billion-hugging-face-deal/", "published_at": "2026-11-02 00:55:39+00:00", "updated_at": "2026-11-02 01:21:58.385402+00:00", "lang": "en", "topics": ["ai-policy", "ai-startups", "ai-infrastructure", "ai-chips", "artificial-intelligence"], "entities": ["Nvidia", "Hugging Face", "Justice Department", "Jensen Huang", "Groq", "SchedMD", "Elizabeth Warren", "Richard Blumenthal"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/justice-department-questions-nvidia-s-12-93-billion-hugging-face-deal", "markdown": "https://wpnews.pro/news/justice-department-questions-nvidia-s-12-93-billion-hugging-face-deal.md", "text": "https://wpnews.pro/news/justice-department-questions-nvidia-s-12-93-billion-hugging-face-deal.txt", "jsonld": "https://wpnews.pro/news/justice-department-questions-nvidia-s-12-93-billion-hugging-face-deal.jsonld"}}