OpenAI has offered the federal government a 5 percent equity stake, worth roughly $42.6 billion at the company’s March valuation of $852 billion, in what would be the largest voluntary transfer of private technology equity to Washington in the modern era. Sam Altman carried the proposal personally to President Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Scott Bessent, according to the Financial Times, with Bloomberg independently confirming the meetings.

The vehicle Altman has in mind is a sovereign wealth fund modelled on the Alaska Permanent Fund, an idea OpenAI floated publicly in an April policy paper calling for a “Public Wealth Fund.” Altman first pitched the concept in early 2025. The structure OpenAI is proposing would obligate rival laboratories, including Anthropic, Google, and Meta, to contribute an identical 5 percent slice. None of them have signed on. A person familiar with the matter told CNBC the White House and Anthropic haven’t discussed a stake at all.

TechCrunch, citing sources close to the talks, described the proposal’s purpose bluntly: to “secure good relations with the administration and address political blowback.” That blowback is real and bipartisan. Senator Bernie Sanders has argued the federal share across the sector should be 50 percent, not five.

The precedent everyone is watching is Intel. Last August the government committed $8.9 billion for a 10 percent position in Intel common stock, and by May Trump was saying publicly he should’ve asked for more. Government ownership, he told reporters, “almost becomes a partnership with the American public.” IBM sits further back in the same lineage.

The mechanics are the hard part. All five outlets reporting the story note that congressional approval would almost certainly be required, which TechCrunch noted “would significantly complicate the matter.” OpenAI filed confidential draft IPO paperwork with the Securities and Exchange Commission in June, though some advisers are reportedly weighing a delay to 2027. The regulatory pressure isn’t abstract: last month OpenAI held back the full public launch of GPT-5.6 at the government’s request, and Anthropic only this week restored foreign-national access to its Fable 5 and Mythos 5 models after a national-security-driven suspension.

What Altman is proposing, then, is less a gift than a settlement. The frontier labs already operate under informal state supervision. Equity would make that arrangement legible, priced, and permanent.

Sources