OpenAI Floats 5% Equity Stake for US Public Fund

A Stake in the Intelligence Economy

Sam Altman has put a concrete number on what sharing AI wealth with the public might look like: a 5% slice of OpenAI, handed to a sovereign wealth fund that doesn't yet exist. The proposal, confirmed by two people with knowledge of the discussions, envisions other AI companies following suit, according to the Financial Times. The move is framed as both a gesture of public partnership and a shield against mounting political pressure as regulators, lawmakers, and the White House circle the industry with growing scrutiny.

At DailyTechWire, we've tracked how capital flows in AI have increasingly become a matter of national strategy, from export controls on Nvidia's H100 chips to the scramble for compute access across Southeast Asia and the Middle East. What's striking here is the bluntness: equity, not tax revenue or licensing fees, as the entry fee for operating at frontier scale in the United States.

Why Equity, Why Now

The timing isn't coincidental. President Trump confirmed in June that he had discussed arrangements where "the American public essentially becomes a partner with the companies," though no percentage was specified then. The 5% figure reported now gives shape to what had been vague talk of public participation. It also aligns with OpenAI's April policy paper, titled "Industrial Policy for the Intelligence Age," which advocated for a public wealth fund that could invest directly in AI labs and deployment firms. That document argued that returns from such a fund "could be distributed directly to citizens, allowing more people to participate directly in the upside of AI-driven growth, regardless of their starting wealth or access to capital."

The proposal is both ideological and tactical. Ideologically, it taps into a strain of thinking, popular in some circles of effective altruism and industrial policy, that sees AI as a national resource warranting collective ownership. Tactically, it's a play to secure regulatory goodwill at a moment when the administration is considering everything from compute allocation rules to new oversight bodies for frontier models.

The phrase "secure good relations with the administration" and "address political blowback," attributed to sources familiar with the discussions, suggests the equity offer is as much about access and influence as it is about redistribution. OpenAI has spent the past year navigating a thicket of governance challenges, from board upheaval to questions about its for-profit conversion structure. A public equity stake could function as both reputational armor and a seat at the table when policy gets written.

The Legislative Gauntlet

Any formal implementation would require congressional approval, a threshold that introduces significant uncertainty. The talks remain preliminary, and the path from boardroom conversation to enacted policy is long and littered with competing interests. Congress has shown little appetite for sweeping AI legislation, and the idea of mandating equity donations, or even accepting them under a new fund structure, would require bipartisan consensus in a fractured political environment.

Senator Bernie Sanders introduced a far more aggressive version of this concept in June: the American AI Sovereign Wealth Fund Act, which would impose a one-time 50% tax on the stock of "systemically important" AI companies. The collected equity would flow into a public fund, with proceeds eventually distributed to citizens. The bill casts a wide net, covering not just model developers but data center operators, infrastructure providers, and robotics firms. Companies like Google or SpaceX, which include AI as one component of a broader business, could spin off non-AI divisions to escape the tax.

That bill has yet to advance to committee, and few expect it to gain traction in its current form. But it signals the kind of policy environment OpenAI is navigating: one where voluntary equity donations may be preferable to mandatory expropriation down the line. By putting a 5% stake on the table now, Altman is shaping the terms of debate before Sanders' 50% becomes the baseline ask.

Valuation, Governance, and the Precedent Problem

A 5% equity stake in OpenAI is worth billions, depending on the valuation snapshot. The company was last reported to be valued north of $80 billion in private markets, which would put a 5% slice above $4 billion. But valuations in AI are volatile, driven by revenue multiples that reflect future inference dominance rather than current cash flow. A stake transferred today could be worth twice as much, or half, within 18 months, depending on competition from Anthropic, Google, and open-weight alternatives, as well as shifts in enterprise adoption and compute costs.

Governance is the thornier issue. Who controls the voting rights attached to that equity? If the fund is structured as a passive vehicle, OpenAI retains operational autonomy but faces public scrutiny over returns. If the fund takes an active role, board seats and strategic decisions become politicized. The document from April suggests OpenAI envisions the fund as an investor, not a regulator, but the line between the two blurs quickly when the investor is a sovereign entity with policy goals beyond return maximization.

There's also the precedent problem. If OpenAI donates 5%, does that become the industry standard? Anthropic, Google DeepMind, and Microsoft's AI division would face pressure to match or exceed the gesture. Startups raising Series A rounds might be asked to set aside equity for public funds as a condition of regulatory approval. The policy, if adopted, wouldn't just redistribute wealth; it would reshape cap tables and investor expectations across the sector.

Asia's Parallel Experiments

While Washington debates equity stakes, governments across Asia are already embedding themselves in AI value chains through direct investment and infrastructure partnerships. Singapore's Temasek has taken positions in several regional AI companies; South Korea's sovereign fund has committed billions to semiconductor and AI R&D; and China's state-backed funds are woven into the fabric of its AI ecosystem, from Baidu to Alibaba Cloud. The U.S. proposal, by contrast, comes late and conditional, a response to the realization that laissez-faire policy has left the public with little claim on one of the most lucrative technology transitions in decades.

The difference in approach reflects divergent philosophies. In Seoul and Singapore, the state is co-investor and co-builder, often from the earliest stages. In the U.S., the conversation begins after companies are already dominant, and the mechanism is donation rather than partnership. That creates a dynamic where equity stakes function more as tribute than collaboration, a toll paid for continued autonomy rather than a shared project from the outset.

What Happens Next

The proposal is still in the realm of negotiation and trial balloon. OpenAI has floated the idea; the administration has expressed interest; Congress has not weighed in with anything resembling consensus. The next steps depend on whether the White House formalizes the concept into a legislative ask, and whether other AI companies signal willingness to participate. If Anthropic, Meta, or Google publicly endorse a similar framework, momentum builds. If they balk, OpenAI's offer becomes an isolated gesture, easy to ignore.

For now, the 5% figure serves as a benchmark. It's specific enough to be credible, modest enough to avoid existential alarm, and large enough to matter in dollar terms. It also sets a ceiling: if OpenAI is offering 5%, Sanders' 50% looks extreme, and any compromise likely lands somewhere in between. The conversation has shifted from whether AI companies should contribute equity to how much and under what terms.

At DailyTechWire, we see this as part of a broader reckoning over who captures the returns from intelligence infrastructure, the same dynamic playing out in debates over compute access, model export restrictions, and the geographic distribution of inference workloads. The equity proposal is one answer, imperfect and incomplete, but indicative of a political climate where doing nothing is no longer a viable strategy for frontier labs. Whether it leads to a sovereign fund, a legislative mandate, or simply a footnote in policy history depends on how the next six months unfold and whether Congress can turn preliminary talks into law.