It is rumored that senior Trump administration officials are in preliminary discussions with major AI companies such as OpenAI: whether the federal government can hold part of the shares of these companies so that the benefits from AI growth can flow back to the public in some way. According to a report by political news website Notus on June 4, this idea was promoted by OpenAI CEO Sam Altman.
People familiar with the matter said Altman raised the idea directly with Trump as early as early 2025 and continued discussions with senior administration officials in recent weeks. There is no agreement, shareholding ratio or legal path at this stage.
In response, Reuters said it could not independently confirm the Notus report, and the White House, OpenAI and Anthropic did not immediately respond to Reuters' request for comment.
Altman wants the government to become the entrance to the distribution of AI benefits
Notus quoted three people familiar with the matter as saying that the relevant discussions centered on "the AI company voluntarily transfers its shares to the government." The investment returns can then be used for public purposes, such as issuing some form of dividend to American households.
This is not a government bailout in the traditional sense. During the 2008 financial crisis, the U.S. Treasury Department took stakes in banks and automobile companies, but that was an emergency arrangement after the companies fell into crisis. OpenAI is not currently in an existential crisis. The logic proposed by Altman is closer to "letting the public hold part of the benefits of AI growth."
A policy document released by OpenAI in April this year also mentioned the idea of establishing a "Public Wealth Fund" so that ordinary citizens who do not have access to the financial market can also share in AI-driven economic growth. Notus reported that OpenAI pointed to the document in response to the interview.
The most direct question is: If AI companies are listed in the future and their valuations continue to rise, will the benefits only go to minority shareholders, or will they be returned to society through public finance, dividends or other public services?
The government serves as both a shareholder and a regulator, and conflicts are difficult to avoid.
The government's holding of shares in AI companies sounds like it would allow the public to share in the technology dividends, but governance issues will immediately become complicated.
If the federal government holds shares in OpenAI, will it just passively collect the proceeds, or will it have voting rights, board seats, and even influence model release, safety standards, and business direction? If the government regulates AI safety while also hoping that the valuation of the AI companies it holds will increase, regulatory standards may be affected by conflicts of interest.
Nat Purser, senior advocate for AI policy at Public Knowledge, told Notus that the problem is that governments will become shareholders and regulators at the same time, which creates a substantial conflict of interest. Jennifer Huddleston, senior fellow for technology policy at the Cato Institute, also worries that government investment will turn into "winner picking," undermining private enterprise and free market principles.
Questions about competition have not been resolved either. Notus reported that OpenAI and Anthropic are preparing for what could be one of the largest initial public offerings in history; however, a person familiar with the matter also said that Anthropic has not discussed providing equity to the government. If only some companies end up accepting government ownership, other AI companies will question whether the government is giving certain companies special status.
Both Trump and Sanders want AI companies to hand over more profits
This discussion is not just a unilateral trial by OpenAI. Notus said that since Trump’s second term, the U.S. government has shown greater interest in holding corporate shares. The U.S. government has directly invested in at least 10 companies, including Intel. Trump previously publicly talked about the Intel deal and said he hoped there would be more similar cases in the future.
When it comes to AI, this thinking even cuts across party lines. Independent Senator Bernie Sanders this week called on the U.S. government to acquire 50% of the equity of AI companies, and plans to introduce a bill to impose a 50% tax on the stocks of AI companies such as OpenAI, Anthropic, and xAI, with the proceeds going into a public sovereign wealth fund.
However, the plan currently discussed by OpenAI and the Trump administration is not close to the “50% equity” mentioned by Sanders. The Notus report emphasized that negotiations are still in their early stages, details are still changing, and no deal may be reached in the end. The legal mechanism is also unclear: how a private AI company transfers its equity to the federal government, what account the government uses to hold it, and how the proceeds are distributed are all yet to be answered.
If government shareholding becomes a reality, American families may participate in AI wealth distribution as "public shareholders" for the first time; but if it becomes more difficult for the government to strictly regulate AI companies, the risks borne by the public will also increase. How to divide AI dividends is changing from a slogan to a system design that must clearly define rights, responsibilities and conflicts.