The Trump administration is reportedly exploring a new model for how the United States participates in the AI boom—one that could involve taking an equity stake in OpenAI or structuring deals so that Americans share directly in the company’s upside.
President Donald Trump, speaking publicly about ongoing discussions, framed the effort around a simple idea: AI success shouldn’t be something that benefits only the investors and executives who happen to be closest to the technology. Instead, he said he is looking at “deals where the American people can benefit from the success of AI.” While the statement is broad and stops short of outlining specific terms, it signals a shift in tone from the more common policy approach of the last few years—regulation, procurement, and incentives—toward something closer to ownership, partnership, and revenue-sharing.
If the administration pursues an equity stake, it would represent a notable escalation in the government’s role in the AI sector. Governments have long supported research and development, but equity is different. It implies not just oversight or funding, but a claim on value creation—along with the political and legal complexity that comes with being a shareholder in a fast-moving, globally competitive company.
What makes this moment especially consequential is that AI is no longer merely a research frontier. It has become infrastructure. The models powering chatbots, coding assistants, customer service automation, and enterprise analytics are increasingly treated like utilities—something businesses build on, governments depend on, and consumers interact with daily. In that context, the question of who captures the economic returns becomes unavoidable. Equity is one way to answer it.
Still, the devil would be in the details. An equity stake could take many forms, ranging from direct ownership to structured financial arrangements that mimic equity exposure without requiring the government to become a traditional shareholder. It could also be paired with governance rights, licensing terms, or commitments around access, safety, and deployment. Each option carries different implications for competition policy, national security, and the future relationship between public institutions and private AI labs.
A “benefit the American people” framing: why equity is on the table
The phrase Trump used—“deals where the American people can benefit”—is telling because it suggests the administration is thinking beyond abstract promises of jobs or general economic growth. Equity is a concrete mechanism. It ties public benefit to measurable financial outcomes rather than relying solely on indirect trickle-down effects.
There are at least three reasons equity-like structures are gaining attention in AI policy circles:
First, AI value capture is concentrating. The most valuable assets in modern AI are not factories or physical supply chains; they are data pipelines, compute access, model training capabilities, distribution channels, and proprietary research. Those assets tend to be controlled by a small number of companies and their backers. If the U.S. wants to ensure that the public shares in the upside, it may look for a direct stake in the entities controlling those assets.
Second, the government already behaves like a strategic investor in other domains. Defense, space, and certain energy transitions have historically involved public-private partnerships where the state takes on risk or provides capital in exchange for long-term leverage. Equity is simply a more explicit version of that logic.
Third, AI is tied to national competitiveness. The U.S. has a strong interest in ensuring that critical AI capabilities remain aligned with American priorities—whether that means safety standards, deployment in key sectors, or resilience against foreign competitors. Ownership or quasi-ownership could be viewed as a tool to influence outcomes without relying entirely on regulation after the fact.
But equity also raises immediate questions: What does the government actually want to own? A minority stake in a company like OpenAI would likely be less about control and more about participation in returns. Yet even minority stakes can come with board observation rights, information access, and negotiation leverage. That alone could reshape how OpenAI and similar firms think about future fundraising, partnerships, and governance.
How an equity stake could work in practice
If the administration is serious about an equity position, it would likely need to navigate a complex landscape of corporate structure, investor expectations, and legal constraints. OpenAI’s ownership and governance arrangements have evolved over time, and any deal would have to fit within existing agreements with current investors and partners.
Several plausible structures could be considered:
1) Direct equity purchase
The simplest concept is the U.S. buying shares—either newly issued shares or secondary purchases. This would require valuation negotiations and would likely trigger scrutiny from market participants concerned about fairness, favoritism, or distortions to competition.
2) Convertible instruments or warrants
Instead of buying equity outright, the government could invest through convertible notes or warrants that convert into equity under certain conditions. This approach can reduce upfront valuation disputes and align the government’s payoff with future performance.
3) Revenue-sharing or profit-participation agreements
Equity is not the only way to share upside. The government could negotiate a contract that entitles it to a percentage of revenues, profits, or licensing fees tied to AI deployments. This might be easier to structure than true ownership while still delivering a “benefit from success” outcome.
4) Licensing and access commitments with financial upside
Another possibility is a hybrid deal: the government receives preferential access to models or deployment rights, while the company receives capital or regulatory support. In return, the government could receive royalties or equity-like upside if certain milestones are met.
5) Public investment funds or sovereign-style vehicles
Rather than a direct government agency holding shares, the administration could route investments through a dedicated fund designed to operate with commercial discipline. This could help manage political optics and reduce the perception of direct political control.
Each structure would change the balance of power. Direct equity tends to create the strongest alignment with financial outcomes, but it also creates the most governance and conflict-of-interest questions. Revenue-sharing can be more targeted and less intrusive, but it may be harder to define in a way that accurately reflects “success” in a rapidly evolving AI business.
The unique challenge: AI success is not one thing
One reason equity deals are tricky in AI is that “success” can mean different things depending on the business model. OpenAI’s value could come from consumer subscriptions, enterprise licensing, API usage, partnerships with cloud providers, or future platform strategies. It could also come from entirely new products that don’t exist yet.
If the administration wants Americans to benefit from AI success, it needs a definition of success that is both fair and enforceable. That means negotiating metrics that survive changes in product strategy. For example, if the company pivots from one revenue stream to another, does the government’s benefit track the pivot? If the company restructures, does the agreement remain intact? If the company’s intellectual property is licensed rather than sold, how is value measured?
These are not academic concerns. AI companies often evolve quickly, and contracts that are too rigid can become meaningless—or worse, create incentives to game the metrics.
A unique take on the policy logic: equity as a “social contract” for frontier tech
There’s a deeper philosophical shift embedded in the equity idea. For decades, frontier technologies have been treated as a cycle: the private sector innovates, the public sector regulates, and society benefits indirectly through productivity gains and new industries. But AI is different in two ways.
First, AI is both a tool and a competitor. It can automate tasks across industries, potentially displacing workers and reshaping labor markets. Second, AI is a platform. The companies that control the models can become gatekeepers to downstream innovation.
In that environment, “benefit the American people” cannot be only about growth. It has to be about distribution and leverage. Equity is one way to formalize a social contract: if the public helps enable the ecosystem—through research funding, compute infrastructure, talent pipelines, and regulatory frameworks—then the public should share in the upside when the ecosystem produces extraordinary returns.
This is not a radical concept in other sectors. When governments invest in infrastructure, they often seek returns through taxes, royalties, or long-term economic development. Equity is simply a more direct mechanism for capturing value when the asset is intangible and concentrated.
Yet there is a tension. Equity can also be perceived as politicizing innovation. If AI labs believe that government involvement will increase unpredictability—through shifting political priorities or future renegotiations—they may hesitate to take risks. The administration would need to design the deal in a way that is stable, legally robust, and insulated from day-to-day politics.
That stability is crucial because AI progress depends on long-term research horizons. Investors tolerate uncertainty when they trust that rules won’t change midstream. A well-structured equity arrangement could provide that trust. A poorly structured one could undermine it.
National security and industrial policy: why OpenAI specifically?
OpenAI is not just another AI startup. It has become one of the most visible and influential labs in the world, with models that shape how millions of people interact with AI and how businesses deploy it. Its prominence makes it a natural candidate for any policy initiative aimed at ensuring American participation in AI value creation.
But there’s also a national security angle. AI capabilities are increasingly tied to defense, intelligence, cyber operations, and strategic communications. Even if the government doesn’t want to control the company, it may want leverage to ensure that critical capabilities are developed responsibly and that access is managed appropriately.
An equity stake could be framed as a way to align incentives: the government benefits financially when the company succeeds, while the company benefits from a stable partnership with the state. In theory, that alignment could encourage responsible deployment and reduce the risk of sudden capability shifts that create security concerns.
However, national security concerns cut both ways. If the government holds equity, it may face pressure to demand transparency or impose constraints. That could collide with the company’s need to protect trade secrets and maintain competitive advantage. The best deals would likely include carefully negotiated boundaries around what information is shared and what remains confidential.
The political calculus: why now?
The timing matters. AI policy has been dominated recently by debates over regulation, safety standards, and liability. Those conversations are important, but they often move slowly compared to the pace