The U.S. government taking a stake in OpenAI is not a hypothetical anymore. Talks between the White House and OpenAI's Sam Altman have been running quietly for over a year, with Altman personally pitching the idea to President Trump in early 2025. This week, Trump publicly confirmed the discussions, telling reporters on Air Force One that Americans could "essentially become a partner" in AI companies.
The structure being discussed is this: OpenAI donates equity to the federal government, which uses it to seed a "Public Wealth Fund." That fund would hold shares, invest in long-term assets, and potentially pay out returns directly to American citizens. Think of it as something like Alaska's oil dividend, applied to artificial intelligence. No equity percentage has been agreed. No legal structure exists yet. Both sides acknowledge the details could change entirely.
So why is this happening now? OpenAI is heading toward a stock market listing later this year, targeting a valuation above $850 billion. The company is generating around $2 billion in monthly revenue. But it is also burning through money at a pace that makes that valuation hard to justify: internal projections show a $14 billion loss in 2026 alone, and analysts estimate the company may need over $200 billion in additional funding to reach profitability by 2030. A government stake before the IPO would give OpenAI something no private investor can offer: a politically connected, institutionally powerful backer with every reason to want the company to succeed.
That last point is also the core problem. When a government owns shares in a company, it has a financial reason to protect that company's value. The same government is also the regulator setting the rules for AI. Those two roles sit in direct tension. A policy advocate at Public Knowledge put it plainly: the government should not be "less willing to impose, or enforce, safety rules because doing so could reduce the value of its own investment." Free-market voices at the Cato Institute have raised similar concerns, arguing that government equity stakes in private firms threaten the basic principles of a competitive market.
This is also not OpenAI's first time building government relationships. The Trump administration has already taken equity stakes in at least ten companies during this term, including a 10% stake in Intel. The administration also recently moved into quantum computing, purchasing roughly $2 billion in equity across nine firms. The OpenAI talks fit a clear pattern: the White House treating strategic technology companies the way past administrations treated infrastructure.
For everyone watching from outside the AI industry, the practical implications are straightforward. If this deal closes, the U.S. government becomes a direct stakeholder in the company that makes ChatGPT. Other major AI companies would face immediate pressure to offer similar arrangements or explain why they have not. Anthropic, which has also proposed public fund concepts but is not currently in equity talks with the administration, filed confidentially for its own IPO just this week. The race to go public is intensifying, and the political stakes are rising alongside the financial ones.
Senator Bernie Sanders has proposed going much further, with a bill that would force major AI companies to hand over 50% of their equity to a federal fund, giving the government board seats and voting rights. That proposal is unlikely to pass. But the fact that both a left-wing senator and a right-wing president are circling the same basic idea, that AI profits should be shared more broadly, tells you something real about where public sentiment is heading. Companies in any industry that interact with AI tools, data, or automation should pay attention: the question of who owns AI's upside is becoming a political issue, not just a financial one.