“The Biggest AI Infrastructure Project in History”
It was huge news last month. Nvidia and OpenAI signed a letter of intent for a strategic partnership that will deploy at least 10 gigawatts of AI data center capacity requiring millions of chips, fast-computing graphics processing units (GPUs), Nvidia’s bread and butter. To support this infrastructure buildout, Nvidia will invest up to $100 billion in OpenAI, progressively as each gigawatt is deployed. The first phase is scheduled to go online in the second half of 2026 using Nvidia’s Vera Rubin platform.
The electricity required for data centers capable of training and running the most advanced AI models rivals national energy consumption. Ten gigawatts can keep the lights on in about 7.5 million homes for a full year. Nvidia’s investment in AI capacity is not just about chips, though. It’s about building the digital equivalent of an industrial power grid.
The investment serves several strategic purposes for Nvidia:
1) Securing Long-Term Demand: Nvidia CEO Jensen Huang has stated that each gigawatt of AI data center capacity could generate $50 billion in revenue (Forbes). By aligning with OpenAI, Nvidia ensures its GPUs remain central to the development of frontier AI models.
2) Preempting Rivals: OpenAI has explored designing its own chips, but with Nvidia’s investment that seems unlikely. Jacob Bourne, technology analyst at eMarketer, told Reuters that the deal “throws cold water on the idea that rival chipmakers or in-house silicon from the Big Tech platforms are anywhere close to disrupting Nvidia’s lead.”
3) Roadmap Integration: Nvidia and OpenAI will co-optimize their hardware and software roadmaps. Matt Britzman of Hargreaves Lansdown writes that the deal “ensures Nvidia’s GPUs remain the backbone of next-gen AI infrastructure” and “cements Nvidia’s leadership position at a time when custom chips from tech giants and startups have started to raise some questions.”
4) Diversification of OpenAI’s Partnerships: The deal may also mean OpenAI will be able to rely less on Microsoft, a $13 billion investor. Bourne said it “signals greater independence as [OpenAI] continues diversifying away from its Microsoft partnership.”
Antitrust Implications
Nvidia’s dominant position in the data center GPU market—estimated at greater than 90%—has already attracted scrutiny from U.S. regulators. Patrick Moorhead, senior contributor at Forbes, wrote that “targeting Nvidia for antitrust investigation was probably inevitable” given its market share and influence over AI infrastructure.
According to Moorehead’s reporting, the Department of Justice is examining whether Nvidia engages in exclusionary practices, such as:
1) Vendor Lock-In: Offering better pricing to customers who use Nvidia chips exclusively.
2) Tying and Retaliation: Requiring purchases of certain products to access others, or penalizing customers who use competitors’ chips.
3) Platform Access Restrictions: Potentially limiting competitors’ ability to interface with Nvidia’s CUDA software platform.
DOJ Antitrust Division Chief Gail Slater recently stated that enforcement must focus on “preventing exclusionary conduct over the resources that are needed to build competitive AI systems and products.” She emphasized that enforcers are examining how dominant firms may foreclose access to key inputs and distribution channels. (Slater also recently lamented the challenges the government faces in enforcing antitrust laws when it comes to technology and law firm Goliaths. Read our post.)
In a keynote address at the Fordham Competition Law Institute conference, Slater underscored the need for vigilant enforcement in the AI sector, saying the focus must be on “preventing exclusionary conduct over the resources that are needed to build competitive AI systems and products.”
Rebecca Haw Allensworth, antitrust professor at Vanderbilt Law School, warned that Nvidia’s financial interest in OpenAI could distort market behavior. “They’re financially interested in each other’s success. That creates an incentive for Nvidia to not sell chips to, or not sell chips on the same terms to, other competitors of OpenAI.”
This concern is amplified by Nvidia’s dominant position in the GPU market, which powers most AI data centers. Preferential pricing or delivery schedules could undermine competition and innovation.
Broader Industry Context
The Nvidia-OpenAI partnership reflects a broader acceleration in AI infrastructure investment. Paulo Carvão, a Harvard Senior Fellow writing for Forbes, described the deal as “the biggest AI infrastructure project in history,” adding that the partnership “aims to ease access to the massive computing power needed to train advanced models, one of the biggest constraints in AI today.” Carvão also noted the tension between rapid AI advancement and the need for global safeguards. Technologists and policymakers are calling for binding international agreements to regulate high-risk AI applications.
In another Forbes article, Kolawole Samuel Adebayo wrote that while “enterprises grumble about uneven ROI from their AI pilots” and “startup founders whisper about inflated valuations,” the underlying investment trend is resilient. In the first half of 2025, 64% of global startup funding value was driven by AI deals, totaling $162.8 billion, according to Forbes.
He described the current investment climate as one of “delayed payoff and accelerated capital,” where investors are betting on long-term infrastructure and enterprise solutions rather than short-term consumer wins. He cited Glilot Capital Partners’ $500 million fundraise as an example of this strategy, quoting co-founder Kobi Samboursky: “Some segments of AI might be overheated, but our strategy is purposefully targeted. Cybersecurity is a durable, mission-critical category where breakthrough technology continues to create outsized returns.” One of Israel’s leading venture capital firms, Glilot’s assets under management have now surpassed $1 billion.
Tech industry insider and popular commentator Kara Swisher has dedicated much of her attention to the competitive dynamics and concentration risks in the AI sector. In a CNN interview discussing Elon Musk’s unsolicited $100 billion bid for OpenAI, Swisher was skeptical about such a deal. In light of market consolidation, she said, “It’s unlikely that regulators would allow a deal like that to go through without serious scrutiny.” Swisher has repeatedly warned that dominant tech firms may suppress competition in AI, similar to what happened in search and social media. “There was no innovation in search after Google won. The same thing could happen in AI if we don’t keep the market open,” she said.
She also pointed to the massive capital flows into AI as a sign of both opportunity and risk. According to Bloomberg, generative AI could produce $1.8 trillion in annual revenue by 2032, representing 16% of all tech spending. Swisher has cautioned that such projections may encourage monopolistic behavior unless agencies intervene.
The announcement really excited investors, though. Nvidia’s market capitalization rose to approximately $4.5 trillion upon news of the investment, reinforcing its position as the world’s most valuable semiconductor company.
Kyla Scanlon is a popular economics commentator on social and traditional media. “For OpenAI [the investment] secures it the compute power [CEO Sam Altman] keeps saying is the bottleneck,” she said on a recent Instagram clip. She characterized the partnership as “a closed circuit of value creation,” summarizing it this way: “OpenAI and other players need massive amounts of compute. The best chips come from Nvidia. Nvidia invests billions of dollars into OpenAI. That cash then funds the purchase of more Nvidia chips. Investors see demand locked in and push Nvidia’s valuation even higher.”
Scanlon cautioned that the price of electricity is climbing in the U.S. “So,” she asks, “are we prioritizing the speculative AI buildout over grid capability for Americans?”
Government Scrutiny and Antitrust Enforcement in AI
As these commentators have observed, Nvidia’s investment in OpenAI solidifies its position at the core of AI infrastructure by locking in long-term demand for its GPUs, directly shaping the development of next-generation AI models, and embedding its technology in the sector’s backbone.
Nvidia and OpenAI are now poised to co-optimize hardware and software, making Nvidia’s systems indispensable for advanced AI development and, in the process, limiting opportunities for rivals. Such vertical integration heightens concerns that industry consolidation could stifle innovation, as just a few companies continue to dominate the AI landscape.
The heightened competitive pressures within the AI industry have not gone unnoticed by government enforcers. Authorities are increasingly concerned with the ways in which leading companies may leverage their dominance over essential AI infrastructure—including compute power, access to data, and distribution channels—to strengthen their market position and suppress competition.
Recent statements by DOJ’s chief antitrust enforcer highlight a significant shift in enforcement priorities from consumer-facing applications to the foundational inputs of AI systems. This reflects a growing commitment among agencies to examine, and if necessary, intervene in transactions and business practices that could impede competition or new entrants.
Vigilance becomes more urgent as AI becomes a critical driver of economic growth and national competitiveness. High-profile deals — such as Nvidia’s investment in OpenAI, along with numerous other transactions documented in the Mogin Law A.I. Deal Table — demand careful review.
Edited by Tom Hagy, Editor-in-Chief of the Mogin Law Blog.
