Washington Weighs an Equity Stake in AI: Inside OpenAI’s Push for a National “AI Dividend”

For most of the last three years, the biggest financial story in artificial intelligence has been about how much private money is flowing into a handful of frontier labs. This week, that story flipped. Instead of asking investors for capital, OpenAI is reportedly proposing to hand a slice of itself to the American public — and the government along with it.

According to people familiar with the discussions, OpenAI has floated giving the United States government an equity stake of roughly 5%, a position that would be worth approximately $42.6 billion based on the company’s most recent private-market valuation of $852 billion. Sam Altman is said to have raised the concept directly with President Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Scott Bessent, positioning it not as a bailout or a regulatory concession, but as a deliberate architecture for how the gains from advanced AI should be shared.

From a Single Company to an Industry-Wide Fund

What makes the proposal notable isn’t just that OpenAI is willing to give up a piece of itself. It’s the scale of what Altman is reportedly imagining beyond his own company. The idea, as described, would have every leading US AI developer — not just OpenAI — contribute roughly 5% of its equity into a single public vehicle. That fund would then be managed in a manner modeled explicitly on the Alaska Permanent Fund, the sovereign wealth structure that invests Alaska’s oil revenues and pays residents an annual dividend simply for being residents of the state.

The comparison is deliberate and rhetorically powerful. Alaska’s fund turned a finite, extractive resource — oil beneath state land — into a renewable stream of shared prosperity. Applying that same logic to AI implies a premise: that frontier models trained substantially on public data, public infrastructure, and public research investment represent a kind of collective resource, and that the wealth generated by commercializing that resource should flow back to the public that helped create the conditions for it, not solely to shareholders and founders.

Why Now?

The proposal did not emerge in a vacuum. It lands at a moment when the AI industry is simultaneously more profitable and more politically exposed than at any point in its short history. Valuations across the sector have detached meaningfully from current revenue, drawing comparisons to the dot-com era from serious financial analysts. At the same time, public anxiety about AI’s effect on jobs, wages, and the concentration of technological power has become a mainstream political issue rather than a fringe concern.

Handing the government an equity stake serves several purposes at once. It gives policymakers a direct financial interest in the success of the company they are also expected to regulate — a dynamic that critics will inevitably call a conflict of interest, and that supporters will call an alignment of incentives. It also provides a politically resonant answer to the accusation that AI wealth is being hoarded by a small number of founders, employees, and venture investors in San Francisco.

The Uncomfortable Distribution Question

Underneath the headline figure sits a much thornier question that the AI industry has mostly avoided answering directly: who actually benefits from the AI boom, and who is being left out of it? A recent UNESCO analysis quantifying the gap in AI access across South Asia offered a blunt reminder that the technology is not distributing its benefits evenly, even as its economic value concentrates in a handful of companies clustered in a few countries.

That unevenness is precisely the gap a dividend-style structure is meant to address, at least domestically. But it also raises hard design questions that have not been publicly resolved. Would payments go to all US citizens, or only to residents of certain states? Would the fund be managed by an independent board insulated from political pressure, the way Alaska’s is, or would it be more directly controlled by the Treasury? Would foreign nationals working at these companies, or foreign investors who supplied much of the capital that built them, have any claim on the structure at all? None of these questions have been answered publicly, and skeptics argue that the framing — AI will generate so much wealth that the public deserves a direct slice of it — is doing a lot of work to paper over how difficult the actual mechanics would be to build.

A Strategic Backdrop: The Multipolar AI Race

It is difficult to separate this proposal from the broader geopolitical context in which it landed. The same week the equity-stake discussions surfaced, Shanghai hosted its 2026 World Artificial Intelligence Conference, with President Xi Jinping attending in person for the first time since the event began in 2018. The optics were hard to miss: while American AI companies debated how to distribute the proceeds of a still-unrealized boom, China’s political leadership was making a personal, public show of backing its own AI ecosystem.

Framed against that backdrop, OpenAI’s proposal reads as much like industrial policy as it does philanthropy. A public equity stake gives the US government a financial reason to champion, fund, and defend its domestic AI champions in a way that a purely regulatory relationship does not. It also gives OpenAI a powerful new argument in any future conflict with regulators: that the government’s own financial interest is now bound up with the company’s success, not just its compliance.

Not Without Precedent, But Unprecedented in Scale

Governments taking equity stakes in strategically important private companies is not new — sovereign wealth funds and state-backed investment vehicles do this routinely in sectors from energy to telecommunications. What would be unusual here is the scale and the sector: a coordinated, industry-wide equity contribution from the most valuable technology companies on earth, structured explicitly as a public dividend rather than a strategic investment stake managed for financial return alone.

If it happens, it would mark one of the more significant shifts in how advanced technology companies relate to the state since the postwar era of large government-directed industrial projects. If it doesn’t — and there are many reasons it might not, from antitrust concerns about coordinated equity contributions across supposed competitors, to constitutional questions about how such a fund would be authorized — it will still have served as a signal of where the most powerful people in AI think the political winds are blowing.

What Happens Next

For now, the proposal remains exactly that: a proposal, reportedly discussed in private meetings rather than filed as formal policy. Whether it advances will likely depend less on the technical merits of the Alaska Permanent Fund comparison and more on the ordinary mechanics of Washington politics — who champions it, who opposes it, and whether it can survive contact with a Congress that has shown little appetite for fast-moving, consequential AI legislation of any kind so far.

What is clear is that the conversation has shifted. A year ago, the dominant AI policy debate centered on safety testing requirements and export controls on advanced chips. Today, a leading AI lab is publicly discussing how to share its equity with the taxpayers whose infrastructure, research funding, and, in some tellings, whose data helped make it possible. Whether or not this specific proposal survives in its current form, it is unlikely to be the last time an AI company treats public equity distribution as a serious policy lever rather than a talking point.

How Other Labs Might Respond

One open question hovering over the entire proposal is whether other frontier labs would actually agree to participate in a shared vehicle, even if OpenAI leads the way. Google DeepMind, Anthropic, and xAI operate under different ownership structures, different investor expectations, and, in Anthropic’s case, a public-benefit corporate charter that already builds broader stakeholder obligations into its governance. Persuading each of these companies to voluntarily hand over a meaningful slice of equity to a public vehicle would require a level of industry coordination that has rarely, if ever, been achieved voluntarily in a sector this competitive.

There is also the practical matter of investor consent. Existing shareholders in these companies, including large venture funds and strategic partners who have poured tens of billions of dollars into the sector, would need to agree to dilution on a scale that a 5% contribution implies. Some may see the political goodwill generated by participation as worth the cost; others may view it as an unnecessary giveaway that weakens their own return profile without a clear governance mechanism ensuring the fund is managed prudently over the long term.

A Test Case for AI Governance More Broadly

Whatever happens to this specific proposal, it is likely to become a reference point in the broader conversation about how the gains from transformative technology should be distributed. Advocates for the idea point out that previous waves of technological disruption — from industrialization to the internet — largely left the question of distribution to market forces and, occasionally, after-the-fact taxation, with mixed results for the workers and communities most affected. A dividend mechanism built into the industry’s growth from the outset would represent a genuinely different approach: distribution as a designed feature of the system, rather than a correction applied years later through the tax code.

Skeptics counter that a voluntary equity contribution from a handful of companies, however large, is a poor substitute for the kind of durable policy framework that only legislation can provide, and that treating a negotiated arrangement between a few CEOs and a few cabinet officials as a substitute for public deliberation sets an uncomfortable precedent of its own. That tension — between the speed and flexibility of a privately negotiated deal and the legitimacy and durability of an actual law — is likely to shape how this story develops in the months ahead, regardless of which specific numbers end up in any final agreement.

Topiry will continue tracking this story as details of any formal proposal emerge, including how a public AI dividend vehicle might be structured, funded, and governed.

About the Author

Topiry

Administrator

Topiry

Visit Website View All Posts