Sam Altman is proposing to give Washington a seat at the table — and a very expensive one at that. According to a Financial Times report citing two people familiar with the discussions, the OpenAI government stake being floated would hand the US administration 5% of the company, a slice worth approximately $42.6 billion based on OpenAI’s $852 billion valuation from its March 2026 funding round. It’s one of the most audacious dealmaking moves in Silicon Valley history, and it tells you everything about how the relationship between Big Tech and Washington has shifted in the past year.
- Sam Altman is pitching an OpenAI government stake of 5%, worth roughly $42.6 billion at current valuation.
- The proposed OpenAI government stake would follow a sovereign wealth model similar to Alaska’s Permanent Fund.
- Altman wants Anthropic, Google, and Meta to offer similar 5% stakes — none have signalled any interest yet.
- The deal would be the first time Washington holds equity in a private AI company, preceding a likely IPO.
Table of Contents
The Pitch: Democratising AI’s Upside
Altman’s framing is characteristically ambitious. He’s positioning the OpenAI government stake not as a regulatory concession or a political favour, but as a mechanism to make sure ordinary Americans actually benefit from the AI boom — not just the venture capitalists and tech insiders who got in early. The proposed structure borrows from the Alaska Permanent Fund, a state-owned vehicle established in 1976 that takes surplus oil revenues and pays annual dividends directly to Alaskan residents. The analogy is deliberate: AI, in this framing, is the new oil, and the government should be extracting value from it on behalf of the public.
It’s a clever narrative. Whether it holds up under scrutiny is another question entirely. Critics will point out that an OpenAI government stake doesn’t automatically translate into dividends for working Americans — the mechanics of how returns would flow to citizens are far from settled, and any such arrangement would almost certainly require Congressional approval before it could go anywhere near a payout structure.
Who Altman Is Talking To
The OpenAI government stake proposal has been raised directly with President Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Scott Bessent. That’s a remarkably senior audience for what the FT characterises as still a conceptual, early-stage discussion. The fact that Altman is having these conversations at the Cabinet level at all signals how seriously both sides are taking the possibility of a formal equity relationship between the US government and its most valuable AI company.
Altman has also reportedly indicated he’d like to see Anthropic, Google, and Meta contribute comparable 5% stakes through the same sovereign vehicle. None of those companies have publicly signalled any appetite for that kind of arrangement. Anthropic in particular has had a fractious few months with the administration — more on that below. Asking a competitor that just emerged from a government-imposed export lockdown to voluntarily hand over equity is, to put it diplomatically, a tough sell.
The Broader Context: Government Equity Is Now a Policy Tool
What makes the OpenAI government stake proposal feel less like a wild idea and more like a logical next step is how aggressively the Trump administration has already been using equity as a lever with the tech sector. Last August, the government took a 9.9% stake in Intel, converting $8.9 billion in CHIPS Act grants into shares priced at $20.47 each. That position is now worth well over $50 billion — a fact that clearly hasn’t been lost on anyone in the White House. Trump himself said in May that he should have negotiated a bigger cut of Intel. AMD and Nvidia, meanwhile, agreed to surrender 15% of their China chip revenues to secure export licences. The pattern is unmistakable: the administration sees equity stakes and revenue-sharing as the preferred instrument for managing its relationships with frontier tech companies.
Against that backdrop, an OpenAI government stake starts to look less like a generous offer and more like a strategic pre-emption. If the government is going to extract value from AI companies one way or another, it’s arguably better for OpenAI to define the terms upfront than to find itself on the receiving end of something less favourable — like, say, Senator Bernie Sanders’ competing bill, which would require the largest AI companies to hand over 50% of their equity to a public fund, with proceeds flowing directly to Americans as cash payments. Fifty percent is a very different number to five.
A Difficult Month for Frontier AI
The timing of these discussions isn’t coincidental. June has been a turbulent month for frontier AI releases in the US. OpenAI launched GPT-5.6 in a restricted form after the White House’s Office of the National Cyber Director asked for a limited rollout while officials build out a testing framework for powerful AI models. That intervention alone would have been significant. But Anthropic had it worse: the company spent most of the month under emergency export controls on its Mythos 5 and Fable 5 models after the Defense Department labelled Anthropic a ‘supply chain risk.’ Access was only restored this week.
OpenAI has consistently positioned itself as a more cooperative partner to the government than Anthropic — signing partnerships that Anthropic declined. The OpenAI government stake proposal fits that pattern perfectly. It’s Altman doubling down on the idea that working with Washington, rather than butting heads with it, is how you protect your ability to operate at the frontier.
What This Means for OpenAI’s IPO
There’s another layer here that’s easy to miss. Both OpenAI and Anthropic have filed confidentially for IPOs. Any OpenAI government stake agreed now would be structured before the dilution that comes with a public float — meaning the government would be getting in at private company terms, then watching its position shift as shares hit the market. That’s a significant financial consideration for both sides of the negotiation.
OpenAI is also currently facing a probe from a coalition of 42 state attorneys general, adding to the regulatory pressure the company is navigating simultaneously. A formal OpenAI government stake — one that gives Washington a direct financial interest in OpenAI’s success — would create an interesting dynamic in that context. It’s hard to be the government’s regulator and its investment partner at the same time.
If the deal does materialise, it would mark the first time the US government has taken an equity stake in a private AI company. That’s a genuinely significant milestone, and not just symbolically. It would signal that Washington sees AI companies not merely as entities to be regulated, but as strategic national assets to be co-owned. The implications for how future AI policy gets made — and who it serves — would be profound, and very much worth watching.
Source: Decrypt
