On January 21, 2025, President Trump unveiled what he called "the biggest AI infrastructure project by far in history"—the Stargate Project, a $500 billion joint venture uniting six titans of artificial intelligence: OpenAI, Oracle, SoftBank, Microsoft, Nvidia, and Arm, plus Abu Dhabi's MGX investment group.
Standing beside Larry Ellison, Sam Altman, and Masayoshi Son in the White House Roosevelt Room, Trump celebrated the "monumental undertaking" as proof of American technological supremacy. The executives gushed appropriately. Son declared it "the beginning of the golden age in America." Altman called it "the most important project of this era," turning to Trump to add: "We wouldn't have been able to do this without you."
What nobody mentioned that evening—and what Congress has barely questioned since—is whether this massive collaboration between competitors violates more than a century of antitrust law designed to prevent exactly this kind of market concentration.
Only one voice has raised serious legal concerns: Madhavi Singh, a researcher at Yale Law School and deputy director of Yale's Thurman Arnold Project, an initiative dedicated to antitrust issues. Her forthcoming paper in the Berkeley Technology Law Journal argues bluntly that Stargate represents "the latest and most flagrant example of the government's enabling private sector companies to expand and entrench their power under the guise of protecting American tech supremacy."
What makes Stargate legally problematic isn't the scale of investment—AI requires massive capital. It's that six industry leaders who compete across multiple product categories have united to form a single company.
As Fortune's Shawn Tully frames it: "Wasn't this something like allowing GM, Ford, Toyota, suppliers Bosch and Lear, and auto software provider Continental AG, to collaborate on building immense car factories?"
The Stargate partnership encompasses:
These aren't complementary businesses filling different market niches. They're competitors who frequently battle for the same customers in overlapping markets.
Singh identifies where these overlaps create competition concerns. Oracle has historically pressured Microsoft and other hyperscalers by charging lower prices and offering flat-fee structures. "Oracle has been a disruptive force in the market," Singh states in her paper. "Now it may adopt Microsoft's pricing strategy. That would raise prices and lower options for customers. Stargate risks elimination of a maverick."
Microsoft, Google, and Amazon are all developing custom GPUs and CPUs to escape Nvidia and Arm's strongholds. "Because of Stargate," Singh writes, "Microsoft might stop challenging Arm and Nvidia in chips."
Stargate has disclosed little about ownership shares, governance, or partner roles. What's known comes primarily from OpenAI's initial press release describing it as a "new company" with four shareholders: SoftBank, OpenAI, Oracle, and MGX of Abu Dhabi.
The scale is staggering. One facility in Abilene, Texas covers roughly the size of Manhattan's Central Park with 1.2 gigawatts of power capacity—enough to power 1 million homes. By September, OpenAI announced a $300 billion-plus program where Oracle will furnish 4.5 gigawatts of capacity for three data centers in Texas, New Mexico, and Michigan.
All told, the projects encompass $400 billion in AI infrastructure and seven gigawatts of capacity—sufficient to power half the households in Georgia. OpenAI stated it should hit the $500 billion commitment goal by the close of 2025.
That half-trillion figure represents roughly twice the cost of Saudi Arabia's King Abdullah Economic City and the International Space Station combined.
Singh grounds her legal analysis in the two primary antitrust statutes: the Clayton Act and the Sherman Act.
"The Clayton Act states that a court will block a joint venture if it shows probable harm to future competition," Singh explains. "It doesn't have to be showing harm yet, but potential damage of loss from head-to-head competition, such as higher prices, reduced choice, and reduced innovation."
The key phrase: "potential harm to future competition." The law doesn't require proof of current damage—merely credible risk that consolidation will eliminate competitive dynamics down the road.
Singh points to the 2021 FTC decision blocking a proposed merger between Arm and Nvidia as precedent. Even though the companies didn't make the same types of chips at the time, regulators blocked the union because it would erase a potential competitor. Each might eventually enter the other's market, providing more choices and lower prices.
Stargate creates precisely this problem across multiple market segments simultaneously.
The Sherman Act bans agreements "in restraint of trade" and prohibits arrangements that deprive markets of independent decision-making centers.
"Now, Arm, Nvidia, and Microsoft separately decide what types of chips they produce," Singh asserts. "The Sherman Act prohibits activity that deprives the market of independent centers of decision-making and therefore the diversity of economic interests. This is precisely what Stargate does."
When competitors coordinate through joint ventures, they lose independence in strategic decisions. Will Oracle continue undercutting Microsoft's cloud pricing when both benefit from Stargate's success? Will Microsoft keep developing competing GPU technology when it shares Stargate facilities with Nvidia?
The temptation to coordinate rather than compete becomes overwhelming when companies share ownership in a massive joint venture.
Singh provides a primer on the "AI stack," which consists of three layers:
She argues competition remains healthy in models and applications—evidenced by DeepSeek and other challengers mounting credible competition to established players. The menace to competition concentrates at the infrastructure layer.
Three players—Amazon, Google, and Microsoft—control 70% of cloud services. Nvidia holds between 80% and 95% of the GPU market. TSMC accounts for 60% of chip production.
Infrastructure is precisely where Stargate operates and where it shrinks an already lean competitive field by putting several of the few players on the same team.
Singh's core argument is that Stargate creates irresistible incentives for cartelization—the division of markets and coordination of pricing that antitrust law explicitly prohibits.
"All of these tech markets seem initially competitive," she told Fortune in an interview. "But it takes a bit of time for anticompetitive barriers to get erected. A lot of these players realized that instead of competing in each other's markets, it makes economic sense to earn monopoly profits, and give each other a share by giving out such favors as IP licenses that are really designed to reward competitors for staying in their own lanes."
The economic logic is straightforward: "I'll take the monopoly in one kind of chips, and you take the monopoly in IP for those or another kind of chips."
Coordination maximizes collective profitability at the expense of customers who face higher prices, reduced choice, and slower innovation.
What's remarkable about Stargate isn't just its structure—it's the near-total absence of regulatory scrutiny or congressional concern.
At a Senate Commerce, Science, and Transportation Committee hearing on May 8 titled "Winning the AI Race," Altman and Senator Ted Cruz extolled the Stargate model. Not a single senator questioned its legality. The term "antitrust" appears nowhere in the three-and-a-half-hour transcript.
Congress's reaction ranges from high praise (particularly from Cruz, whose state of Texas hosts Stargate's biggest facilities) to near-total silence from everyone else. Regulators have issued no challenges.
Singh argues the Trump administration has veered from rigorous antitrust enforcement toward light-touch regulation prioritizing two objectives: supporting U.S. AI giants as "national champions in the U.S.-China trade war" and allegedly protecting national security by enabling these companies to collaborate domestically.
"The joint venture has clear federal backing, which makes it unlikely the federal authorities would investigate it," Singh told Fortune.
The highest-profile criticism came from Elon Musk—notably absent from the Stargate partnership. The day after Trump's announcement, Musk trashed the deal on X, charging the group "doesn't have the money" to fund Stargate. He reposted an image of a crack pipe alongside allegations that Altman and associates were "freebasing" to generate the $500 billion figure.
Those attacks infuriated Trump's staff and marked the start of tensions that led to Musk's May departure from DOGE.
But as Fortune notes, "Musk effectively did the Stargate founders a favor by not mentioning the real threat they pose to AI's progress." He focused on financing rather than anticompetitive structure—missing the legal vulnerability Singh identified.
Defenders might argue that Stargate partners continue competing vigorously. Hyperscalers are developing proprietary GPUs to escape Nvidia's pricing. OpenAI entered data centers to reduce dependence on Microsoft. Nvidia courts "neoclouds" like CoreWeave to lessen reliance on dominant hyperscalers.
Surface-level competition persists—for now.
Singh's concern is that coordination incentives will prove irresistible once Stargate operations mature. "It makes economic sense to earn monopoly profits, and give each other a share" rather than compete head-to-head across market segments.
History suggests companies acting through joint ventures eventually coordinate rather than compete, particularly when coordination maximizes collective returns.
The implicit defense of Stargate is national security and competitiveness versus China. If U.S. AI companies need to collaborate to compete globally, perhaps antitrust enforcement should accommodate that strategic imperative.
Singh rejects this reasoning. Protecting "national champions" doesn't require abandoning competition law domestically. And the benefits of domestic competition—innovation, efficiency, consumer welfare—contribute more to long-term competitiveness than protecting incumbent market positions.
"If the Trump paradigm would make a few protected players far richer, and minimize the payoff for our citizens and producers," Fortune concludes, "it's a bad deal for America."
Singh's Stargate analysis raises a larger question: Do the equity investments, purchase agreements, and partnership arrangements tying together AI rivals warrant scrutiny under competition laws?
Microsoft's $13 billion investment in OpenAI, Amazon's investment in Anthropic, Google's relationships across the ecosystem—the entire AI industry is characterized by cross-investments and partnerships unusual in other sectors.
Perhaps these arrangements enable necessary collaboration in capital-intensive infrastructure development. Or perhaps they're gradually cartelizing the industry, with Stargate representing the most explicit and aggressive example.
Singh is "virtually a lone voice challenging Stargate as a downer for competition," Fortune notes. But lone voices sometimes prove prophetic—particularly when legal analysis is sound and regulatory capture prevents broader scrutiny.
The question isn't whether Stargate violates the letter of antitrust law. Singh makes a credible case that it does. The question is whether anyone with enforcement authority cares enough to challenge it.
So far, the answer is no.
If your organization is navigating strategic partnerships, joint ventures, or collaboration agreements and needs guidance on antitrust implications and competitive positioning, Winsome Marketing's team can help you structure arrangements that enable cooperation without crossing legal lines.