OpenAI floats giving the US government a 5% equity stake

TL;DR:

• OpenAI is in early-stage talks to give the US government a 5% stake, with chief executive Sam Altman arguing public ownership is the fairest way to spread AI’s gains.
• The proposal, first reported by the Financial Times, would ask other US developers to hand Washington similar equity via a fund modelled on Alaska’s oil-wealth vehicle.
• The discussions are described as “conceptual” and could require an act of Congress.

OpenAI has discussed giving the US government a shareholding of around 5%, according to the Financial Times, as the industry courts an administration that is tightening its grip on advanced AI. Altman has framed public ownership as the best way to let ordinary Americans share in the technology’s rewards, and has raised the idea with President Trump, commerce secretary Howard Lutnick and treasury secretary Scott Bessent.

A stake for other labs too

The structure floated by OpenAI executives would ask each of the largest US developers to allot 5% of their equity to an investment vehicle modelled on the Alaska Permanent Fund, the state corporation that turns oil revenue into dividends for residents. Reuters reports it is not yet clear whether rivals such as Anthropic, Google or Meta would agree. Both OpenAI and Anthropic have previously raised the idea of public wealth funds and “digital dividends” in policy papers, and the talks land as the pair prepare stock-market listings that some investors think could value each above $1tn (£751bn).

The proposal follows a fortnight of Washington muscle-flexing over frontier models. The Commerce Department this week lifted its export curbs on Anthropic’s Fable and Mythos models, weeks after imposing them, while officials edge towards voluntary standards for how new models are released. A government shareholding would add OpenAI to a lengthening list: Washington took stakes in Intel and MP Materials last year by converting federal support into equity.

For UK readers, the sovereignty angle is the real story. Forrester analyst Indranil Bandyopadhyay warned that a pre-IPO government stake might calm US investor nerves about regulation but would prompt “other jurisdictions to demand analogous arrangements”, pushing European and Asia-Pacific buyers to reassess data-sovereignty and neutrality assumptions about American providers. That is the awkward question for British firms building on US models: once Washington is a part-owner of the labs, claims of neutrality become harder to sustain.

Looking forward

The talks are early and, by the FT’s account, may need congressional sign-off, so nothing is imminent. Yet the direction is striking. After a period in which the UK and EU argued over how hard to regulate the labs, the US is now weighing part-ownership of them instead. For Britain — leaning on its AI Safety Institute and voluntary commitments — a state-backed American AI sector would reshape the competitive and sovereignty calculus well beyond Washington.