Trump calls for investment in U.S. AI companies, with talks with OpenAI, Anthropic, and xAI as soon as this week; Altman throws out the idea of a “public wealth fund.”

Trump recently announced that the U.S. government might take stakes in America's top AI companies, calling it "making them partners in the revolution." He is expected to meet with leaders of these tech giants at the White House as early as this week to discuss.

(Background: U.S. lawmakers propose: Public ownership of half of OpenAI's shares, and a 50% stock tax on AI giants)

(Additional context: Anthropic's "Mythos Preview" model outperforms human experts in decision-making, with a success rate of up to 64%)

Table of Contents

Toggle

• From White Paper to White House
• What does government equity stake mean?
• Different paths, same destination

The largest leading AI companies in the market are actively planning IPOs this year, and the Trump administration now seems eager to get a piece of the action, proclaiming "the wealth generated by AI should return to the American public," rather than all remaining in the pockets of a few Silicon Valley shareholders.

Earlier, Trump told reporters aboard Air Force One, "The leaders of all major AI companies are expected to come to the White House next week to discuss the idea of government stakes in these companies." He said:

It almost becomes a partnership with the American public...

The American people can benefit from the success of AI, so they will prefer it.

From White Paper to White House

According to CNBC's report on Friday, OpenAI CEO Sam Altman has been pushing this idea since early 2025, and an insider revealed that discussions have been ongoing for over a year.

OpenAI also released a policy white paper in April this year, officially naming this concept the "Public Wealth Fund." The document states that fund profits can be "directly distributed to citizens, allowing more people—regardless of their starting wealth or access to capital—to participate directly in the upside of AI-driven growth."

In simple terms, the government would hold shares in OpenAI and distribute the profits from these shares to American citizens in some form—similar to a sovereign wealth fund, but with beneficiaries being individuals instead of the state.

This idea is not without precedent. Last year, the Trump administration acquired a 10% stake in Intel as a rescue measure for the struggling chipmaker. Subsequently, the government also took stakes in IBM, quantum computing companies, and key mineral enterprises. From this perspective, investing in OpenAI is not an unprecedented red line-crossing but an extension of existing policy logic. However, the valuation of this time’s target is higher, and the strategic implications are more complex.

Currently, no official investment terms have been finalized; details may still change.

What does government equity stake mean?

The development logic of top AI labs over the past decade has been built on high-risk private capital investments: venture capital bears uncertainty in exchange for future excess returns. If government stakes break this cycle, it could subtly alter the decision-making chain regarding research directions, model deployment, and data usage. Not necessarily through commands, but through subtle alignment of interests.

A more fundamental question is: Can AI companies maintain independence in their technological paths while accepting government capital? Historically, tech companies with government stakes often have to consider additional stakeholders’ opinions on sensitive decisions.

From Trump’s perspective, this move addresses two issues simultaneously: demonstrating a political commitment to voters that "AI wealth is shared with the public," and securing an official position in the AI arms race. But at what cost remains unclear.

Different paths, same destination

Another noteworthy point is that Senator Sanders earlier proposed a one-time 50% stock tax on OpenAI, Anthropic, and xAI. His logic differs from Trump’s language, but the goal is similar: "giving the public a direct role in deciding the future of this technology," ensuring that the trillions of dollars generated by AI are used to improve everyone’s lives.

One route involves government holding shares in exchange for public profit-sharing; the other involves directly taxing AI companies’ stocks. The approaches are quite different, but both focus on the same issue: the distribution of AI wealth can no longer be avoided.

Critics’ concerns are also straightforward. A former Microsoft employee, Dare Obasanjo, wrote on Bluesky: "This is paving the way for government rescue of OpenAI." The premise is that if OpenAI faces financial difficulties in the future, government stakes could obligate taxpayers to bail it out.