OpenAI held talks with the Trump administration about giving the U.S. government a five percent stake in the company. The Financial Times reported this, citing two people familiar with the discussions. According to them, the talks are still at a very preliminary stage. However, OpenAI CEO Sam Altman presented the plan directly to President Donald Trump and other White House officials. A five percent stake in OpenAI would be worth approximately $42 billion. That is based on the March funding round that valued the company at $852 billion.
Altman's offer
Altman argues that if the company gives the public a financial stake, people will benefit from the rise of artificial intelligence. And it is not just about OpenAI. The proposal envisages other major U.S. companies in the sector, namely Anthropic, Google, and Meta, giving the government a similar stake. These stakes would be pooled in a kind of sovereign wealth fund that would manage assets growing alongside the expansion of artificial intelligence.
Altman chose the Alaska Permanent Fund as a model. Alaska established it in 1976 to deposit surplus oil revenues. The fund invests in stocks and pays an annual dividend to state residents. At the end of May, it was worth nearly $91 billion. However, it is unclear whether the other companies would join the initiative. A source familiar with the matter told CNBC that the Trump administration and Anthropic have not yet discussed any government stake.
According to CNBC, a possible government stake has been under discussion for more than a year, and Altman first proposed the idea to the White House in early 2025. Commerce Secretary Howard Lutnick and Treasury Secretary Scott Bessent have also participated in the talks.
In April, OpenAI published a document outlining a new approach to dealing with the impact of so-called superintelligence, meaning artificial intelligence that surpasses even the brightest people. At the time, the company proposed creating a public wealth fund that would ensure every citizen receives a share of the economic growth generated by artificial intelligence, including those who do not personally invest in the stock market.
Trump himself spoke about similar ideas last month. He told reporters that he had heard proposals under which part of these companies could belong to the American public, making people partners in them. He called the idea interesting and promised that the administration would look into it. Senator Bernie Sanders also weighed in. According to the FT, he is pushing for a much broader version of the plan. He wants the government to acquire a full fifty percent stake in all major companies in the sector through a sovereign wealth fund.
White House pressure is behind the offer
The White House continues to decide when OpenAI and Anthropic may offer their most powerful models to a broad audience. Altman himself described that process as less than optimal. If the government held a stake in the company, it would have a direct interest in whether to permit or delay the release of a particular model. Altman therefore proposed another path as well. On Wednesday, he introduced the idea of a U.S.-led international forum that would establish rules for artificial intelligence. This would allow the government to invest without having to interfere so heavily in development.
Not everyone views the offer favorably, however. Investors told Axios that the entire idea comes across as a publicity stunt intended to create the impression that the public will gain something from the artificial intelligence boom, just as the technology is threatening their jobs. One investor who put money into both Anthropic and OpenAI described it outright as a political move to curry favor with the administration.
David Sherman, an artificial intelligence and financial inclusion strategist at io.net, also criticized it. According to him, a government stake would represent a troubling milestone that would harm competition among companies. He said it would give one company the government's seal of approval, while millions of developers, researchers, and businesses would be left out because of rising prices and a shortage of computing power.
The Intel precedent and the question of Congress
The U.S. government has taken similar steps before. Last August, it acquired a stake of nearly ten percent in Intel, and it has paid off so far. The chipmaker's shares have jumped by almost 400 percent since then, although the broader growth of the entire semiconductor sector also contributed. But Intel was a different case. The government acquired that stake through the CHIPS Act. An agreement with artificial intelligence companies would most likely require separate legislation approved by Congress.
There is also the question of what a government stake would actually achieve. Aside from bringing companies in the sector closer to what is currently one of their biggest obstacles: the government itself. Added to this are concerns about whether Washington's rules governing model releases are already weakening the lead of U.S. companies over China.
Other proposals have also emerged for how the public could share in the profits from artificial intelligence. Some advocate a levy on a portion of pre-tax profits, while others propose a tax on every token generated. Artificial intelligence governance adviser Kevin Bankston said on X: just tax them. Bill Gates proposed a robot tax as early as 2017, which could slow down automation. According to him, robots should pay taxes just as employees pay income tax.
Altman himself described his position in a Financial Times opinion piece at the beginning of the week. He wrote that companies develop the technology, but the rules must be created by citizens and their elected representatives.
