Washington is no longer just asking how to regulate artificial intelligence. It is now weighing whether the public should own a piece of the companies building it.
The latest signal from Washington is not another hearing, pledge or voluntary safety framework. Senior U.S. officials have held preliminary discussions with major artificial intelligence companies about the government taking equity stakes in their businesses, according to Reuters, which cited a NOTUS report based on people familiar with the matter.
That is a very different conversation from the one founders have grown used to. For the last two years, the federal AI debate has mostly centered on model safety, export controls, copyright, energy use and whether Congress can write rules quickly enough to matter. Equity changes the tone. It turns the government from referee into potential shareholder.
No deal has been announced. No structure has been confirmed. The talks appear early, and the report said OpenAI CEO Sam Altman has discussed the idea with senior Trump administration officials while Anthropic is not involved in conversations about giving the government equity. OpenAI, Anthropic and the White House did not immediately respond to Reuters requests for comment on the NOTUS reporting.
Even with those caveats, the discussion matters because it points to a broader shift in how Washington views strategic technology. AI is no longer being treated as only a commercial software market. It is being treated as infrastructure, industrial capacity and national leverage in the same breath.
For founders and investors, the first question is simple: what exactly would the government be buying, and on what terms? A passive, non-controlling stake would look very different from a preferred share with special rights, information access or approval powers over future transactions.
That distinction is not academic. AI companies already sit in complicated capital stacks, with cloud commitments, strategic investors, venture funds, employee ownership plans and, in some cases, nonprofit or public benefit structures layered on top. Adding the federal government to that mix would make governance more sensitive and future financing rounds more political.
OpenAI is the obvious name because of its scale, influence and relationship with Washington. But the logic would not stop there. Any company building frontier models, critical AI infrastructure or defense-relevant tools could eventually face the same question. If the government believes a technology is too important to leave fully to private capital, then ownership becomes one more policy instrument.
That would make life more complicated for startups operating in sensitive verticals. A founder selling AI tools into defense, health care, energy, cybersecurity or chips may already deal with procurement rules and export controls. Government equity could add another layer: conflict reviews, disclosure requirements, political scrutiny and questions from customers overseas about whether the company is still purely private.
Washington already has a precedent problem
This is not happening in a vacuum. The Trump administration has already shown a willingness to take stakes in strategic technology companies. The U.S. government took a 9.9 percent ownership stake in Intel in 2025, and the Commerce Department announced in May 2026 that it would provide about $2.013 billion in incentives to nine quantum computing companies while receiving minority, non-controlling equity stakes.
That quantum package included major allocations such as $1 billion for a new IBM quantum chip manufacturing effort and $375 million for GlobalFoundries, with the government receiving about a 1 percent stake in GlobalFoundries. The point was not only financial return. It was industrial policy with ownership attached.
AI would be a sharper test because model companies are not chip foundries. Their value sits in talent, training data, model weights, inference scale, brand trust and enterprise distribution. Those assets can move faster than factories, and their risks are harder to measure. A government stake in a chip company says Washington wants domestic capacity. A government stake in a frontier AI lab says Washington wants influence over the direction of a general-purpose technology.
That is why the global comparison matters. Gulf sovereign wealth funds and Asian state-backed investors have long used capital to secure access to strategic sectors. The U.S. has usually preferred subsidies, procurement and regulation. Moving toward equity stakes would make American industrial policy look more like a sovereign investment model, even if officials present it as voluntary and limited.
There is also a timing issue. President Trump signed an executive order this week directing agencies to seek voluntary early access to advanced AI models. That order sits beside a wider push to test AI systems for national security risks before they are released more broadly. Ownership talks, if they advance, would sit on top of that access agenda.
For entrepreneurs, the practical lesson is not that every AI startup should expect a government term sheet. It is that strategic relevance now changes the funding conversation. The more a startup touches national security, compute infrastructure or frontier model capability, the more likely its cap table becomes part of a policy discussion.
Venture investors will not ignore that. Some may welcome a federal stake as a validation signal, especially if it brings procurement access or political protection. Others will worry about slower exits, foreign buyer restrictions, governance friction and the risk that a change in administration turns yesterday's advantage into tomorrow's liability.
The next thing to watch is structure. If Washington explores non-voting shares, capped returns or public-purpose dividends, the idea may be sold as a way to let taxpayers share in upside from technologies built with public support. If it asks for special rights, board access or operational influence, the market will read it very differently.
AI companies have spent years telling the public that their technology will reshape the economy. Washington seems to be taking them seriously. The question now is whether that seriousness arrives as oversight, capital or a seat at the table.
Also read: Washington tightens the chip loophole around Nvidia’s AI boom • Apple gives Poke a rare lane into iMessage for AI agents • OpenAI makes ChatGPT memory more useful across conversations
