OpenAI Proposes Handing the U.S. Government a 5% Stake Worth $42.6 Billion

In the span of a single week, OpenAI watched Washington delay the public launch of its flagship GPT-5.6 model, then floated an extraordinary olive branch: hand the U.S. government a 5% ownership stake in the company, a slice worth roughly $42.6 billion at OpenAI's current valuation.

The proposal, first reported by the Financial Times and confirmed across multiple outlets on July 2, would see OpenAI donate one-twentieth of its equity to a government-controlled vehicle modeled on the Alaska Permanent Fund — the state trust that pays annual dividends to residents from Alaska's oil wealth. According to two people familiar with the talks, CEO Sam Altman wants every leading American AI developer to contribute the same 5% share, turning a bilateral deal into an industry-wide public wealth fund.

At the $852 billion post-money valuation OpenAI set in its record-breaking March funding round — the largest private financing in history — a 5% stake pencils out to about $42.6 billion. That figure alone would make the federal government one of the most valuable single shareholders in any private company on Earth.

The Backdrop

The timing is impossible to ignore. News of the talks surfaced just six days after OpenAI delayed the full release of GPT-5.6 at the government's request, with Commerce Secretary Howard Lutnick reportedly cautioning Altman against shipping the model without prior approval. Rival Anthropic spent most of June with its Claude Fable 5 and Mythos 5 models disabled worldwide under the first U.S. export controls ever applied to an AI model rather than to hardware.

Against that pressure, the equity offer reads less like altruism than statecraft. CNBC reported the donation is intended to "secure good relations with the administration and address political blowback." Altman is understood to have raised the concept directly with President Donald Trump, Commerce Secretary Lutnick, and Treasury Secretary Scott Bessent, and to have spoken with Senator Bernie Sanders (I-Vt.) in recent weeks. He first pitched a government stake to the administration in early 2025, and OpenAI formally floated a "public wealth fund" in an April policy paper.

Critically, this is a proposal, not a signed deal. The FT characterized the discussions as conceptual and early-stage, and reported that implementing any arrangement "might require an act of Congress." The all-labs version would pull equity from Google, Meta, xAI, and Anthropic — none of which have indicated they would participate. OpenAI declined to comment to the FT, and the White House did not immediately respond.

Why It Matters

A government equity stake in a frontier AI lab would rewrite the relationship between Washington and Silicon Valley. The Trump administration has already run a version of this playbook on chipmakers: last August it converted CHIPS Act grants into a 9.9% stake in Intel at $20.47 per share, and secured agreements from AMD and Nvidia to hand over 15% of their China chip revenue in exchange for export licenses. Extending that logic from semiconductors to the software companies building general-purpose intelligence marks a profound shift toward state co-ownership of the technology many believe will define the century.

The 5% figure is also, notably, the smallest number attached to public ownership of AI to date — and that is the point of contention. Sanders has gone far further, filing the American AI Sovereign Wealth Fund Act in June, which seeks 50% of the voting shares of major U.S. AI firms through a fund his office values at $7 trillion, enough to pay every American a roughly $1,000 annual dividend. In an accompanying op-ed, Sanders argued bluntly: "The Public Should Own Half of the Big A.I. Companies." Vice President JD Vance has said Trump prefers giving the public equity over cash payouts.

That gap — 5% offered versus 50% demanded — frames the coming fight. A voluntary 5% donation buys OpenAI goodwill and a seat at the table while ceding minimal control; a legislated 50% tax would hand the government blocking power over corporate decisions. Both proposals raise thorny questions: Would government shareholders vote their shares? Would a state stake chill competition, entrench incumbents, or create conflicts of interest as the same government both regulates and profits from these firms? And how does public ownership square with the confidential IPO filings both OpenAI and Anthropic have already submitted?

What to Watch

For now, the ball sits with the administration and, ultimately, Congress. Watch whether the Treasury and Commerce departments formalize the concept into a term sheet, and whether any rival lab breaks ranks to signal interest — a single defection could turn Altman's industry-wide vision into reality or expose it as a solo gambit. Watch, too, for movement on Sanders' bill, which sets the aggressive ceiling against which OpenAI's 5% will be judged. And with OpenAI facing a probe from a coalition of 42 state attorneys general and an eventual public listing on the horizon, any government shareholding negotiated now would lock in Washington's position before the ownership expansion a full float brings. The question is no longer whether the U.S. government takes a stake in frontier AI — but how big, and on whose terms.