OpenAI Proposes Giving the US Government a 5% Stake Worth $42 Billion

OpenAI CEO Sam Altman is floating one of the most audacious political proposals in the history of American technology: give the federal government a 5% ownership stake in the company—and push every other major US AI lab to do the same.

The plan, first reported by the Financial Times and confirmed by Bloomberg and CNBC on July 2, 2026, is designed to transform the United States government from a regulator and customer of AI technology into a direct financial beneficiary of it. At OpenAI’s current $852 billion valuation, a 5% stake would be worth roughly $42.6 billion—an extraordinary figure that would give Washington a seat at the table in the most consequential technology race of the century.

”Alaska for AI”

The intellectual architecture of Altman’s proposal draws from an unexpected source: the Alaska Permanent Fund, the state’s sovereign wealth vehicle that distributes a portion of oil revenues to every Alaskan resident as an annual dividend. Altman has reportedly framed his vision as an “Alaska-style” arrangement for AI—one where the public doesn’t just bear the risks of powerful artificial intelligence but shares in its financial rewards.

Under the broader version of the proposal, OpenAI would not be the only company making the offer. Altman wants Anthropic, Google, and Meta Platforms to each grant the government equivalent stakes in their AI operations. Whether those companies would agree is far from certain. Anthropic declined to comment; Google and Meta did not respond to requests before publication.

The specifics of how the stakes would be structured remain undefined. The discussions are described by people familiar with them as conceptual and early-stage. Any formal arrangement would likely require an act of Congress, and the political path to get there is treacherous—particularly in an environment where the administration has repeatedly clashed with AI companies over model access restrictions.

Why Now

The timing of Altman’s overture is not accidental. OpenAI is navigating one of the most complex political moments in its history. The US government’s security review of GPT-5.6—the next-generation model that has been in a holding pattern since late June while federal agencies assess its capabilities—has left Altman in an awkward position: eager to deploy a powerful new system while satisfying national security demands he cannot publicly detail.

The proposed stake is, at least in part, a political sweetener. By inviting the government into OpenAI’s equity structure, Altman signals a willingness to share not just the responsibilities of frontier AI development but its financial fruits. The move would also complicate any effort by Congress to impose punitive regulations: legislators would be voting against their own constituents’ financial interests.

“Giving the public a financial interest in the company is the best way to share the upside of AI,” Altman told associates in conversations described to reporters. The framing echoes language he has used publicly about the need for AI’s benefits to be broadly distributed, while translating it into a concrete—and self-interested—policy mechanism.

The Conversations Behind the Proposal

Altman has raised the idea with a notable cross-section of Washington power. On the executive side, he has spoken with President Donald Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Scott Bessent. The outreach to Bessent—whose Treasury Department oversees sovereign wealth matters—suggests Altman is thinking seriously about how such a fund would be structured and governed.

Perhaps more surprising is Altman’s reported conversation with Senator Bernie Sanders (I-Vt.), a longtime critic of concentrated corporate power and a vocal skeptic of the AI industry’s governance structures. That Altman is courting Sanders suggests he is trying to build the broadest possible coalition: free-market conservatives who can be sold on American technological competitiveness, and progressives who respond to the language of shared prosperity and public ownership.

What Would Happen to the Stake?

The proposal leaves open a fundamental question: what would the government actually do with a 5% stake in OpenAI? Sovereign wealth funds typically invest for financial returns; the US does not currently have a formal federal wealth fund, though the Trump administration has discussed creating one. A government stake in a private AI company would raise novel questions about governance, voting rights, conflict of interest, and the relationship between regulatory oversight and financial ownership.

One scenario floated in discussions is that the stake would flow into a newly created federal AI wealth fund, with proceeds eventually distributed to American citizens—a direct analog to the Alaska model. Another possibility is that it would be held by an existing federal entity, such as the Treasury or a newly chartered government investment corporation.

Either way, the proposal would fundamentally alter the relationship between Silicon Valley’s most powerful AI companies and the US federal government. Critics will argue that giving the government a financial stake in AI companies creates perverse incentives—regulators with equity interests are less likely to impose tough oversight. Proponents counter that the current arrangement, in which taxpayer-funded research and infrastructure subsidizes private AI wealth creation, is already deeply imbalanced.

The Broader Stakes

The proposal arrives at a moment when questions about who benefits from AI—and who bears its risks—are becoming the defining political fault line of the technology. OpenAI’s IPO filing, submitted confidentially in May, revealed a company on a trajectory toward $25 billion in annual revenue. Anthropic recently crossed $1 billion in quarterly revenue. The financial stakes are enormous, and Washington is paying attention.

For Altman, the 5% offer is a strategic gamble. It costs OpenAI something—dilution, a new powerful stakeholder, potential governance complications—but buys something potentially more valuable: political cover, a friendlier regulatory environment, and a narrative in which the government is a partner in AI’s success rather than an adversary trying to contain it.

Whether the gambit works depends on whether Congress can agree on a mechanism to accept and hold such stakes, whether other AI companies will follow, and ultimately whether voters find the “Alaska for AI” framing compelling. None of those outcomes are guaranteed. But in a political environment defined by anxiety about AI’s concentration of wealth and power, Altman is betting that the best defense is a very expensive offense.