An equity offer that runs through Congress is a proposal, not an instrument. That is the receipt the paper carries on OpenAI's pitch to hand the U.S. government a roughly 5 percent stake — worth about $42.6 billion at the company's $852 billion valuation — with other leading American labs asked to contribute similar shares into a public investment vehicle modeled on Alaska's Permanent Fund. [1] The paper's July 7 account of why the government stake still needs Congress named the proposal-not-instrument posture. Today the posture holds, and the gap between the announcement and the enacting law is the whole story.

The pitch, first reported by the Financial Times on July 2, remains conceptual. Sam Altman has raised it with President Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Scott Bessent, and has spoken with Senator Bernie Sanders. [1] The structure would pull equity from Google, Meta, and Anthropic as well — none of which has indicated it would participate — into a sovereign-wealth-style fund that pays out from investment returns, the way Alaska distributes oil wealth to residents. [2] And, per the FT's own framing, implementing any such deal "might require an act of Congress." [1] That conditional is not a footnote. It is the difference between a term sheet and a talking point.

On X, the idea splits three ways. Populist-finance accounts read it as overdue — if AI creates enormous wealth, the public should own the upside, and OpenAI's own framing that the stake would "clear political obstacles" only confirms the transactional logic. [3] Sovereignty hawks read it as nationalization-lite, government fingers in the most powerful private technology of the era. And the left reads 5 percent as far too small: Sanders has floated 50 percent, making OpenAI's offer look less like public ownership than a public-relations hedge. MSM, meanwhile, treats it as a novel public-wealth experiment — the mechanics of an Alaska-model fund, the valuation math, the cast of Cabinet officials in the room. [2]

What the Alaska comparison obscures is the governance question underneath. Alaska's Permanent Fund holds passive, diversified investments and pays residents a dividend; it does not sit on the boards of the companies it owns or vote its shares to shape their conduct. A 5 percent stake in a single frontier lab is a different animal. Equity carries rights — to information, sometimes to a board seat, sometimes to a vote on major decisions. Would the government take those rights, or waive them? A passive stake is a revenue claim; an active one is a lever over the most powerful private technology of the era. The proposal, as floated, does not say which it would be, and that ambiguity is precisely what makes both the populists and the hawks able to read it their own way.

The paper's middle is the missing instrument. No bill has been introduced. No executive framework has been published. No term sheet, no binding structure, no named lab beyond OpenAI has committed. The proposal exists as a conversation, and conversations do not convey equity. The mechanism by which 5 percent of a private company's shares would move into a public fund — whether by legislation creating the vehicle, by an executive arrangement, or by some negotiated grant tied to regulatory forbearance — has not been specified. Each path implies a different set of legal authorities, and none of them has been invoked.

What sharpens it is the balance sheet underneath. The government-stake idea and OpenAI's cash crunch are the same document. The company is floating equity-for-goodwill while simultaneously seeking roughly $100 billion in new funding and projecting about $14 billion in losses this year. The 5 percent offer is being pitched as political-blowback management by a company that needs both capital and cover — and it is being pitched, notably, the same week OpenAI leaned toward delaying its IPO to 2027. An equity stake that would attach to a listing being pushed a year out, and that would require a Congress that has done nothing, is not equity. It is an intention.

The follow-up test is unchanged: has anything converted the pitch into a binding vehicle? A bill, an executive framework, a term sheet, a second lab signing on. Absent any of those, the story is a proposal still failing to become an instrument — floated, discussed, and unratified, while the clock it would attach to keeps slipping.

-- THEO KAPLAN, San Francisco