What OpenAI's 5% government stake would mean in practice

What OpenAI's 5% government stake would mean in practice

Mon, August 17, 2026 at 6:46 PM GMT+3 4 min read

American AI companies are collectively worth trillions of dollars, and lawmakers, Nobel-winning economists, and AI executives themselves are asking whether ordinary citizens should own a piece of them.

Sam Altman, CEO of OpenAI, has been discussing a plan to give the U.S. government a 5% equity stake in his company, according to a Financial Times report from early July. Altman wants other leading AI firms to do the same, with equity flowing into a government-run fund modeled on Alaska's Permanent Fund, which invests oil wealth and distributes annual payments to residents. The talks are in early stages, and any deal would likely require an act of Congress.

The proposal comes as public opinion on AI has turned sharply negative. More than 200 economists, including 16 Nobel laureates, have warned that AI could put millions of workers out of jobs within years, not the decades that previous waves of automation required. A fund could cushion blow, if it's structured properly.

Why Altman's 5% offer is smaller than it sounds

Altman first raised the idea with the Trump administration in 2025, according to CNBC, and spent the next year building his case. OpenAI published a formal proposal in April calling for a "Public Wealth Fund" that would give every citizen, including the tens of millions of Americans who don't own a single share of stock, a stake in AI-driven growth.

Each American's share of the proposed stake would be worth between $125 and $146, depending on whether OpenAI goes public at its target valuation of $1 trillion or stays at its current $852 billion. That's a fraction of what Alaska's Permanent Fund typically pays to the state's more than 600,000 residents. Even in a down year, its payouts are robust: the Permanent Fund delivered $1,000 per resident in 2025, its lowest payout in five years. Because the fund is worth more than $91 billion, having been built over five decades on continuous oil revenue and diversified investments, its dividends are consistent and high enough to mean something to each person.

Altman's version would serve 342 million people with all of its money in a single company that's not yet profitable. OpenAI generated $13.1 billion in revenue in 2025 and spent more than twice that much running the business, according to audited financial statements verified by the Financial Times. Alaska's fund pays out because it earns returns on a diversified portfolio every year. A government fund holding OpenAI stock wouldn't see similar returns anytime soon.

Norway's Government Pension Fund Global, the world's largest sovereign wealth fund, held about $2.2 trillion at the end of 2025 and took half a century to get there, funded continuously by the country's oil revenue. Altman's proposed starting stake would amount to less than 2% of what Norway has built.

Sanders' $7 trillion alternative and what the precedents show

Sen. Bernie Sanders has proposed something far larger. In June, the Vermont independent introduced the American AI Sovereign Wealth Fund Act, which would require AI companies earning more than $200 million a year in AI-related revenue to surrender half their stock to a government fund. A seven-member commission, nominated by the president and confirmed by the Senate, would manage the fund and hold voting shares. Sanders' office estimates it would be worth about $7 trillion, enough to send more than $1,000 to every American through annual dividends. The commission would also have the power to block corporate decisions it deemed harmful to the public.

Altman's plan is voluntary and wouldn't give the government a board seat. Sanders' bill is compulsory and would give a public commission equal board representation and voting power over corporate strategy. Altman's plan depends on rivals agreeing to participate, and so far none have. The Trump administration and Anthropic, one of OpenAI's closest competitors, haven't discussed the government taking stakes in Anthropic, according to the Financial Times, and Google and Meta haven't signaled interest.

The U.S. government has taken equity stakes in private companies before, and the track record shows that ownership without governance accomplishes little. During the 2008 financial crisis, the Treasury Department used the Troubled Asset Relief Program to invest almost $80 billion in the auto industry alone, and by the end of 2012 held about 22% of General Motors. Taxpayers recovered most of the money, but not all of it. More recently, the Trump administration took a 10% stake in Intel, converting $8.9 billion in CHIPS Act grants into common stock with no board seat and no say in Intel's strategy. Intel later warned that the government ownership could complicate international sales.

If the goal is to ensure AI benefits "all of humanity," as OpenAI's own charter states, a 5% stake that gives the public money in the company and no say in how it is run doesn't advance it. A researcher at Convergence Analysis, a technology policy think tank, argued in a July 2025 paper that government equity stakes in AI companies should come with governance rights and transparent mandates. A 5% shareholder, however, would be the smallest owner at the table.