OpenAI’s Latest Idea: Give 5% of the Company to the American Public
The sovereign wealth fund proposal is the most interesting thing OpenAI has done that has nothing to do with models.
On July 2nd, the Financial Times reported that Sam Altman has proposed donating 5% of OpenAI’s equity to a US sovereign wealth fund, and that the proposal calls for other AI companies to make similar contributions. The stated purpose is straightforward in the way that Washington rhetoric usually is: secure good relations with the administration and address political blowback. The subtext is equally clear. The AI industry is facing mounting pressure from a government that both wants the technology to succeed domestically and wants the economic upside distributed more broadly than a handful of shareholders.
The idea is not as new as it sounds. Altman has been publicly discussing a public AI fund for over a year. OpenAI’s policy paper released in April, titled “Industrial Policy for the Intelligence Age,” laid out a framework where a sovereign fund could invest directly in AI labs and the companies deploying their technology. The returns, the paper argued, could be distributed directly to citizens, giving people who do not have access to venture capital or stock options a way to participate in AI-driven economic growth. That is a genuine political argument, and it is one that has bipartisan resonance in a way that most tech policy does not.
The current administration has already weighed in. President Trump confirmed in June that his team had discussed concepts where pieces of AI companies could be given to the American public, with the public becoming a partner in the upside. The language matters here. Calling citizens “partners” rather than consumers or users is a deliberate framing, and it suggests that the political pressure on AI companies to demonstrate broad economic benefit is not going away.
Congress would need to approve any formal structure, which is where the proposal runs into real friction. Tax policy and corporate equity structures are among the most politically complicated areas in Washington. The Sanders bill introduced in June takes a more aggressive approach, proposing a one-time 50% tax on stock held by “systemically important” AI companies, with the collected shares deposited into a public wealth fund. That bill applies to companies dealing with AI data centers, infrastructure, and robotics. It would also allow companies like Google and SpaceX to spin off non-AI portions of their business to avoid the tax. The bill has not advanced to committee, but the fact that it exists as a serious legislative conversation tells you something about the political direction of travel.
The interesting part of the OpenAI proposal compared to the Sanders bill is the voluntary element. Asking companies to donate equity is different from taxing it. The 5% figure is specific enough to matter but small enough that the major AI labs might accept it as an alternative to more aggressive legislative action. The hope in the proposal, presumably, is that a voluntary contribution buys goodwill and regulatory breathing room in a way that fighting a tax bill in Congress does not.
The bigger question is whether the structure OpenAI is proposing actually delivers on its stated goals. A sovereign wealth fund that holds equity in companies that are actively competing with each other creates obvious conflicts. Would the fund vote its shares in OpenAI’s interest or in the interest of the broader AI ecosystem? Would holding equity in both OpenAI and Anthropic, for example, compromise the fund’s ability to assess these companies fairly? These are not rhetorical questions. They are the kinds of structural problems that sink sovereign wealth funds before they start.
There is also the valuation problem. OpenAI is not a public company. Its shares trade in a private market with limited liquidity and prices that reflect negotiated transactions rather than public market discipline. Donating 5% of a private company’s equity to a sovereign fund requires agreeing on what that equity is worth. OpenAI’s last known valuation was $157 billion in a secondary transaction. Whether that number holds up to the kind of scrutiny a sovereign wealth fund would require is a different question.
For the AI industry more broadly, the proposal signals something important. The era of AI companies being able to operate as pure technology plays, indifferent to their political and social effects, is ending. The pressure to demonstrate that AI benefits are broadly distributed is becoming a structural requirement rather than a public relations exercise. Companies that figure out how to navigate that pressure proactively will have more freedom to operate than those that fight it. OpenAI is trying to get ahead of the issue with a proposal that is generous enough to be credible and specific enough to be taken seriously. Whether Congress agrees to the structure or rewrites it into something unrecognizable, the conversation itself is a marker of where the industry is heading.
