The idea, floated in preliminary discussions between OpenAI and the Trump administration, is unlike anything the technology industry has proposed in Washington: hand the government a roughly 5% stake in the company, held for the benefit of the American public. Sam Altman, OpenAI’s chief executive, has argued privately and publicly that letting the public share in the economic gains of AI is the best way to manage the technology’s impact, and he has suggested that the major American AI companies each contribute about 5% of their equity to a public investment vehicle modeled on Alaska’s Permanent Fund.
The proposal is at an early stage, according to people familiar with the discussions, and no agreement has been reached. But the fact that it is on the table at all marks a shift in how the industry’s most prominent company approaches its regulators. For years, OpenAI and its peers prepared for conflict with Washington — lobbying against rules, resisting disclosure requirements, and treating government interest as a threat. The equity proposal represents the opposite instinct: bring the government in as a shareholder with a stake in the company’s success.
The context makes the shift legible. The White House has reportedly asked OpenAI to delay the release of its next model, GPT-5.6, a request that would have been unthinkable during the industry’s earlier period of unchecked release schedules. At the same time, the government has been negotiating voluntary safety standards with the leading AI labs, and it has been weighing export controls that would restrict who can access the most advanced models. OpenAI’s proposal, in this reading, is an attempt to convert an adversarial relationship into a partnership before the relationship defines the company’s fate.
The Alaska Permanent Fund comparison is deliberate. The fund, created in 1976, distributes a share of the state’s oil revenue to every Alaskan resident each year, and it has become a model for how a government can own a stake in an industry without managing it. Altman has cited the fund in conversations with officials, according to people familiar with the matter, as evidence that a government equity stake need not mean government control — the state collects the returns, while the company runs the business.
The mechanics of such an arrangement would be complicated. A 5% stake in OpenAI, valued on the basis of the share prices that private investors have paid, would be worth billions of dollars, and the government would need a legal structure to hold, vote, and eventually sell it. The companies involved would have to agree on how much equity to contribute and when, and the terms would have to be settled in a way that does not trigger the conflict-of-interest rules that govern government ownership of private assets. People familiar with the talks said none of these details have been resolved.
The proposal also has political logic. The administration has made American AI leadership a priority, and it has sought to answer critics who argue that the benefits of AI are accruing to a handful of companies and their shareholders. A public dividend stream, funded by the industry’s success, gives the government a story to tell about AI working for ordinary Americans. For OpenAI, the equity offer buys goodwill at a moment when goodwill is scarce, and it positions the company as the AI lab most willing to accommodate Washington.
The reaction among the other AI companies has been mixed, according to people familiar with the discussions. Some executives see the proposal as a reasonable trade — a slice of equity in exchange for a predictable policy environment — while others worry that a government stake in one company sets a precedent that could spread, and that the government’s presence on a shareholder register could complicate everything from acquisitions to fundraising. None of the companies have publicly endorsed the idea.
Investors, for their part, are watching with a mixture of curiosity and concern. OpenAI’s valuation has been built on the assumption that the company can operate freely in the global market, and any arrangement that ties its equity to government priorities could change how future rounds are priced. At the same time, a government stake could reduce the regulatory risk that has weighed on the entire industry, and some investors have said they would welcome the clarity.
The broader pattern is a renegotiation of the relationship between the AI industry and the state. The industry’s early years were defined by the belief that AI should develop free of government constraint; the current moment is defined by the recognition that the government is too important a customer, regulator, and gatekeeper to be ignored. OpenAI’s equity proposal is the most dramatic example of that recognition — an offer to share the spoils in exchange for a seat at the table.
Whether the government accepts remains an open question. A 5% stake in a private company is a new kind of public asset, and the administration would need to decide how to hold it, manage it, and explain it. But the conversation itself is the story: for the first time, the leading AI company is proposing to treat the U.S. government not as an obstacle to its growth, but as a partner in it.
