FOREMY INSIDER BRIEFING · AI POLICY & WEALTH DISTRIBUTION
A Sovereign Wealth Fund, Built From Equity Instead of Oil
One of the more striking proposals to surface in AI policy circles this summer isn’t about safety testing, chip export controls, or copyright litigation. It’s about ownership. A leading AI lab has floated handing the US government a meaningful equity stake in the company, reportedly worth in the tens of billions of dollars at current valuation, with a pitch that goes well beyond that single company. The broader idea being discussed is an arrangement in which every leading US AI company allocates a slice of its equity, on the order of a few percent, into a public vehicle modeled explicitly on a state sovereign wealth fund that invests natural resource revenue and pays residents an annual dividend.
The pitch, in plain terms, is this: AI is expected to generate enormous wealth over the coming decade, concentrated in a small number of companies and their early investors and employees. Rather than let that wealth accrue entirely to private balance sheets, the proposal would have the public hold a direct, permanent equity claim on the upside, distributed the way a resource-rich state distributes returns from a shared natural asset to its residents.
Why This Framing, and Why Now
The choice of model matters. A sovereign fund built on natural resource revenue works on the premise that a resource belongs, in some meaningful sense, to the public before it belongs to whoever extracts and sells it, and that the extracting company owes the public a cut regardless of who did the digging. Applying that logic to AI is a deliberate rhetorical move: it recasts frontier AI capability itself as something closer to a shared national resource — built partly on public research funding, public infrastructure, and, more contentiously, on data and content produced by the public at large — rather than purely as the product of a private company’s own capital and labor.
That framing lands very differently depending on who is doing the listening. To a policymaker worried about AI concentrating wealth and power in a handful of companies, it offers a mechanism to broaden the benefits without slowing the technology down. To a skeptic of the underlying company’s motives, the timing looks strategic in a less flattering way: proposing to give the public equity is also an effective way to build political goodwill at a moment when that lab may need regulatory patience, favorable treatment on infrastructure permitting, or simply a friendlier climate in Washington as competition and scrutiny both intensify.
The Uncomfortable Questions Underneath the Pitch
Below the surface of an appealing headline, this kind of proposal raises a set of genuinely hard questions that haven’t been resolved publicly:
- Governance. An equity stake without governance rights is a very different thing from one with a seat at the table. Would a public equity stake come with voting power, board representation, or influence over safety and deployment decisions, or would it function purely as a passive financial claim?
- Valuation and dilution. Committing a fixed percentage of equity at a moment of extremely high, arguably speculative, private valuations locks in a specific bet about how that value will hold up. If valuations correct significantly, the public stake shrinks in real terms right along with everyone else’s, without necessarily protecting the public from the downside the way a resource royalty typically does.
- Global equity, not just national. A model built explicitly around one country’s public fund does little for the rest of the world, even though frontier AI companies operate, deploy, and draw training data globally. Recent analysis highlighting how unevenly AI access and infrastructure are distributed across regions, with substantial gaps in some parts of the world, is a pointed reminder that a purely national wealth-sharing mechanism sidesteps the international dimension of the same underlying concern.
- Precedent for other industries. If the public gets an equity stake in AI companies on the theory that the technology is built on shared resources, similar arguments could be made about other industries built substantially on public research funding or public data, which raises the question of why AI specifically is being singled out for this treatment now.
The Political Reception
Reports that the pitch has been made directly to the country’s most senior economic and political officials suggest the lab is treating this as more than a thought experiment, and the involvement of a state-level sovereign fund as the explicit template signals an attempt to borrow legitimacy from a model with decades of public trust behind it, rather than inventing an entirely novel financial structure from scratch. Whether the proposal advances in anything close to its current form will depend heavily on details that haven’t been made public: which specific companies would be asked to participate, whether participation would be voluntary or tied to favorable regulatory treatment, and how a stake in a private, not-yet-public company would actually be valued, held, and eventually distributed to the public in a form resembling a dividend.
It’s also worth noting that this proposal arrives during a year when several leading AI companies have been exploring public listings, with speculation around valuations that could reach into the trillions of dollars for at least one major lab. A public equity carve-out negotiated before any such listing would look very different, financially and legally, than one negotiated after a company is already trading on public markets with existing shareholders and disclosure obligations to navigate.
Why This Matters Beyond the Headline
Strip away the specific mechanics, and the proposal is really a marker in a much larger, unresolved debate: as AI generates historic amounts of wealth, who has a legitimate claim to a share of it, and through what mechanism should that claim be exercised? Government equity stakes are one possible answer. Others being discussed elsewhere in policy circles include taxation specifically targeted at AI-driven productivity gains, licensing regimes for training data that would compensate the people and organizations whose content trained the models, and direct public investment in AI infrastructure that would give governments a stake the old-fashioned way, by simply being an investor rather than a recipient of a negotiated gift of equity.
Each of these mechanisms distributes power differently. An equity stake handed over voluntarily by a company keeps the initiative, and the framing, firmly with the company. A tax or licensing regime imposed by legislation keeps the initiative with government, and is generally much harder for the affected companies to shape to their own advantage. Which path actually gains traction will say a great deal about who holds genuine leverage in this relationship over the next few years, and it’s a fair bet that companies proposing the equity-stake model, on their own terms, understand that distinction perfectly well.
What Ordinary Households Should Actually Expect
It is worth being direct about the gap between the appealing headline and the realistic near-term impact on an ordinary household budget. Even in the most generous version of the proposal currently being discussed, any eventual dividend would depend on the performance of a small number of extremely high-valuation, still-private companies, distributed across an entire national population, arriving only after governance, valuation, and legal structuring questions that have not yet been resolved are actually settled. That is a meaningfully longer and less certain path than the framing of “the public gets a stake in the AI boom” suggests at first read. None of that makes the proposal meaningless, but it does mean the realistic timeline for any household to see a tangible benefit from this specific mechanism is almost certainly measured in years rather than months, and depends on political and legal developments that are, at this stage, still genuinely unresolved. Anyone reading the headline as a promise of an imminent check in the mail should treat it, for now, as a policy proposal worth watching rather than a benefit already secured.
The Foremy Take
A voluntary equity gift, offered on a company’s own terms and timeline, is a fundamentally different animal from a tax or a licensing obligation imposed by law. One keeps the company in control of the narrative and the mechanism; the other doesn’t. Watch closely which model actually gains political traction over the next year — it will tell you a lot about who really holds the leverage in the relationship between AI companies and the public whose resources, data, and goodwill they are increasingly asking to share in.
What to Watch Next
- Whether any formal legislative or executive proposal actually emerges from these preliminary conversations, versus the idea quietly fading from the agenda.
- Whether other leading AI labs publicly respond to, endorse, or distance themselves from the proposal.
- How any eventual public listing by a major AI lab interacts with, or complicates, a pre-negotiated public equity stake.
- Whether alternative wealth-distribution mechanisms, such as data licensing regimes or targeted taxation, gain more traction than the equity-stake model as the policy debate matures.
This report is part of Foremy's ongoing AI Insider Report series, tracking the economics, infrastructure, and policy decisions shaping the AI industry. Foremy Team, foremy.com/.