OpenAI Offers Washington a Piece of Itself: The $42.6 Billion Bid to Buy Political Peace

An unusual offer

Companies spend fortunes lobbying to keep the government out of their business. This week, OpenAI reportedly did the opposite: it offered to hand Washington a piece of itself.

According to the Financial Times, first reported on July 2 and corroborated by Euronews, Engadget, and CNBC, OpenAI has proposed giving the US government a 5% ownership stake in the company. Based on the roughly $852 billion valuation investors placed on OpenAI during a March 2026 funding round, that slice would be worth about $42.6 billion — a genuinely large sum, even by the standards of a company swimming in capital.

The ambition does not stop at OpenAI's own cap table. Per the same reporting, CEO Sam Altman wants other leading American labs — Anthropic, Google, Meta, and xAI among them — to hand over comparable 5% stakes, effectively creating a government-owned sliver of the entire US frontier-AI industry. It is one of the most striking proposals the sector has produced, and it is worth being precise about what is confirmed and what is not.

What is actually confirmed

The core facts trace to the Financial Times, with multiple outlets reproducing the same figures: a 5% stake, roughly $42.6 billion in value, pegged to the March valuation. Altman has reportedly discussed the idea with President Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Bessent, according to Euronews. Trump has acknowledged the talks but has not confirmed any agreement.

Beyond that, much remains soft. Engadget notes the talks are described as early-stage and that any deal would require Congressional approval — a substantial hurdle that the reporting does not gloss over. No structure, no timeline, and no binding commitment from any of the other named companies has been confirmed. Anthropic, Google, Meta, and xAI are named as targets of Altman's pitch, not as participants who have agreed to anything. Readers should treat the "government owns a piece of all of AI" framing as one person's proposal, not an emerging fact.

The Alaska model — and the politics behind it

Altman has been circling this idea for a while. Euronews notes he first floated a version in early 2025 and has previously pitched a "public wealth fund" that would invest in AI companies and distribute profits to ordinary Americans. The explicit reference point is Alaska's oil dividend — the Alaska Permanent Fund, which shares state resource revenue with residents each year. The pitch, in Altman's telling, is that giving the public a direct financial interest in AI is the best way to share the technology's upside.

That is the generous reading. The timing invites a less generous one. As Tom's Hardware points out, the offer surfaced just days after Washington delayed the full public launch of OpenAI's GPT-5.6, having requested early access and additional oversight before broader release. It also lands against a backdrop of a June executive order — described by Engadget as a scaled-back measure — requiring AI companies to share their most powerful models for voluntary government review roughly 30 days before public release, and the earlier episode in which Anthropic was ordered to restrict access to its Mythos- and Fable-class models before being allowed to restore them. In other words, the industry is under real regulatory pressure, and a 5% gift to the government is a conspicuous gesture to make in that climate.

Why a company would give equity away

The logic is not as strange as it first sounds, and there is a fresh precedent. Engadget notes that the Trump administration took a 10% stake in Intel — a stake Trump later claimed had grown substantially in value. Equity has, in effect, become a tool of industrial policy in this administration: the government takes a position, and the company gains a powerful stakeholder with a direct interest in its success.

For OpenAI, the calculation may be that a financially invested government is a friendlier regulator, a more motivated defender against antitrust or nationalization pressure, and a partner rather than an adversary on export controls and model-release rules. A 5% stake is cheap insurance if it buys goodwill on the decisions — model approvals, chip access, liability — that determine whether the company can operate at all. Framing it as a public dividend is smart politics on top of that.

The conflict-of-interest problem

Here is the tension the proposal cannot escape: the same government being asked to regulate frontier AI would also own a piece of the companies it regulates. If Washington holds equity worth tens of billions, its incentive to approve a model release, wave through a data-center buildout, or soften a safety rule is no longer purely about the public interest — it is partly about protecting the value of its own holdings. Several outlets flagged exactly this conflict risk, and it is the strongest argument against the idea.

Nor is everyone convinced the offer is generous enough. Euronews reports that Senator Bernie Sanders dismissed the 5% proposal as insufficient, instead calling for a one-time 50% tax on the shares of major AI companies. That reaction is a useful reminder that "give the government equity" is not an obviously left-wing or right-wing idea — it can be read as public-spirited redistribution or as regulatory capture dressed up as generosity, and different critics will land in very different places.

Hype versus substance

It is worth resisting two overreactions. The first is treating this as a done deal: it is not. It is an early-stage proposal, reliant on a single primary source (the FT), requiring Congressional buy-in, and lacking any confirmed commitment from the other labs Altman named. The "US government to own slice of entire AI industry" headline is aspiration, not policy.

The second overreaction is dismissing it as pure spin. Even as a trial balloon, the proposal signals something real about the moment: the largest AI companies increasingly see their fate as bound to the state — for compute, for chips, for export clearance, for model-release approval — and are willing to trade equity for alignment with it. Whether or not this specific deal happens, that direction of travel is the actual story.

The takeaway

OpenAI's reported offer to give the US government a 5% stake, worth roughly $42.6 billion, is the boldest sign yet that the frontier-AI industry and the American state are fusing their interests. Sold as an Alaska-style public dividend, it doubles as a hedge against regulatory pressure that has already delayed GPT-5.6 and briefly pulled a rival's models offline. The upside — sharing AI's gains with the public — is real in principle. So is the danger: a regulator that owns what it regulates has a harder time saying no. For now, treat it as a serious proposal under negotiation, not a settled arrangement — and watch whether any of the other named labs actually follow OpenAI's lead, because that, more than the 5% figure, is what would turn a headline into a new structure for the industry.