OpenAI has entered early-stage discussions with the Trump administration over a proposal to hand the United States government a roughly 5 percent equity stake in the artificial intelligence company, according to people familiar with the matter. The initiative, framed as a vehicle for public participation in AI’s economic upside, immediately ignited debate over whether such a stake would amount to a novel form of public-private partnership or a dangerous path toward regulatory capture.

The talks remain tentative and informal, but they signal a dramatic new phase in the relationship between Big Tech and Washington. By offering the government a direct financial interest in its success, OpenAI appears to be betting that alignment of incentives will foster a more favorable regulatory environment—even as critics warn that it could undermine the very independence that regulators are meant to preserve.

A Proposal with Few Precedents

Details of the proposed stake are scarce. The 5 percent figure, described by one source as a “rough” number, has not been formalized in any term sheet, and it remains unclear whether the shares would be granted outright, sold at a discount, or tied to specific policy concessions. OpenAI, which was recently valued at over $150 billion following a $6.6 billion funding round, would effectively be diluting existing investors—including Microsoft, which holds a significant minority stake—by issuing new equity to the government.

Under the initial concept, the U.S. Treasury or a newly created federal entity would hold the shares. The government would not receive a board seat or direct operational control, but it would stand to benefit financially as OpenAI’s valuation rises, creating a return that could be funneled back to taxpayers or earmarked for AI safety research. Proponents inside the administration argue that such participation could give the public a direct stake in the technology’s wealth creation, akin to a sovereign wealth fund’s investment in strategic industries.

Why Would OpenAI Offer a Stake?

For OpenAI, the calculus appears to blend defensive and offensive motives. On the defensive side, the company faces an uncertain regulatory landscape as governments worldwide rush to draft new oversight regimes. Federal agencies and lawmakers have scrutinized everything from data privacy to algorithmic bias and the existential risks of advanced AI. A financial partnership could, in theory, soften that scrutiny by making the government a co-investor in success rather than a detached enforcer.

On the offensive side, a government stake could insulate OpenAI from competitors and geopolitical rivals. As China and other nations pour state resources into AI development, an American government-backed champion might be better positioned to compete—a model reminiscent of the early days of the semiconductor industry or the defense-aerospace sector. Moreover, by embedding itself in national economic policy, OpenAI could lock in long-term procurement relationships and shape the technical standards that will govern the industry.

Critics counter that the offer is less a goodwill gesture than a maneuver to co-opt regulators. “When the referee owns a piece of one team, the game ceases to be fair,” said one technology antitrust expert, who requested anonymity because of the sensitivity of the topic. The fear of regulatory capture—where agencies tasked with oversight become champions of the entities they are supposed to police—looms large, especially given OpenAI’s history of governance tensions and its recent embrace of a for-profit structure.

Regulatory Capture: Old Fears in a New Industry

The specter of regulatory capture is not new. From the Interstate Commerce Commission’s cozy relationship with railroads in the 19th century to more recent revolving-door controversies at financial agencies, the risk that regulated industries will come to dominate their overseers is a perennial concern. What makes the OpenAI proposal distinct is its explicit nature: rather than capture happening subtly through lobbying and post-government employment, it would be baked into the corporate structure from the start.

If the government holds equity, every regulatory action that benefits OpenAI’s bottom line—or harms its competitors—could be seen as self-dealing. Would the Federal Trade Commission hesitate to investigate anti-competitive behavior if doing so might tank the value of a publicly held asset? Would the National Institute of Standards and Technology favor OpenAI’s safety protocols over those of rivals, knowing it has a financial stake in the outcome? Even the perception of partiality could erode public trust.

OpenAI has not publicly commented on the talks, but people close to the company say the proposal includes firewalls intended to separate the government’s ownership interest from its regulatory functions. One idea under discussion is to place the shares in a blind trust managed by an independent fiduciary, whose decisions would be walled off from day-to-day policy work. Skeptics note that such arrangements are notoriously leaky in practice, and that the mere existence of a financial link creates perverse incentives.

Sovereign Wealth Fund Comparisons

Some administration officials have likened the potential stake to the operations of a sovereign wealth fund, such as Norway’s Government Pension Fund Global or Singapore’s Temasek Holdings. These state-owned pools invest in a diversified portfolio of assets, including domestic companies, with the aim of generating returns for citizens while insulating investment decisions from political interference. In the AI context, the government would essentially be taking an early-stage equity position in a high-growth technology firm, much as venture capitalists do.

But sovereign wealth funds typically hold passive stakes across many industries, whereas OpenAI would be a concentrated bet on a single, highly volatile sector. Moreover, the federal government is not a typical investor: it writes the rules under which its portfolio companies operate. Norway’s fund, for instance, is managed by Norges Bank Investment Management, which operates largely independently of the government’s regulatory arms. No such separation exists in the U.S. system, where the Treasury Secretary is both a political appointee and the potential custodian of the stake.

Proponents argue that the U.S. already blurs these lines in other areas. The federal government holds patents on life-saving drugs developed with public grants, and it owns a significant portion of the nation’s land and mineral rights. An equity stake in AI, they say, is merely an extension of the principle that taxpayers should benefit from publicly funded research and national strategic assets. Yet these precedents involve tangible property rights rather than ownership of a private corporation, making the analogy imperfect.

Implications for AI Governance

The timing of the discussions is notable. The Trump administration has signaled a deregulatory posture toward technology, with the President vowing to repeal the previous administration’s executive order on AI safety and replace it with a more industry-friendly framework. OpenAI CEO Sam Altman, who has oscillated between calls for strict regulation and warnings against overly burdensome rules, recently praised the administration’s approach as “pragmatic.” A government stake could cement that comity, but at the cost of embedding bias into the regulatory process.

For the broader AI industry, the proposal raises questions about level-playing-field competition. If OpenAI secures a government investor, rivals like Anthropic, Google DeepMind, and startup Cohere might feel compelled to seek similar arrangements, accelerating a government pick-winners dynamic. That could stifle innovation by concentrating resources in a few favored firms and narrowing the diversity of approaches to AI development and safety.

Moreover, the international dimension cannot be ignored. If the U.S. government becomes a direct shareholder in OpenAI, other nations might view the company as an arm of American industrial policy, limiting its access to global markets. Europe, which has aggressively pursued AI regulation through the EU AI Act, might see the stake as evidence that self-regulation by American tech firms is a facade. Trust, already frayed by transatlantic data disputes, could erode further.

Industry and Public Reaction

Initial reaction from the tech policy community has been sharply divided. Some venture capitalists and tech lobbyists applaud the idea as a creative solution to the coordination problem between government and industry. “This is smart statecraft,” said a Silicon Valley investor who has backed several AI unicorns. “If we want the U.S. to lead in AI, the government needs skin in the game.”

Others are less sanguine. Digital rights groups have warned that the proposal effectively monetizes the regulatory process. “This is not public participation; it’s public capture,” said the director of an AI ethics watchdog. “Once the government owns a piece of OpenAI, every public safety measure that hurts the company’s profits becomes a political liability.”

On Capitol Hill, the idea has attracted bipartisan curiosity but also deep skepticism. Some lawmakers see potential in requiring companies that dominate critical infrastructure to share equity with the public, pointing to models like Alaska’s Permanent Fund, which returns oil revenues to citizens. Others worry that the arrangement would concentrate even more power in the executive branch, bypassing Congress’s constitutional authority over federal property and spending.

The Path Forward

For now, the discussions remain exploratory, and no formal proposal has been presented to the National Economic Council or the Treasury Department. The legal and logistical hurdles are formidable: issuing shares to the government could trigger securities law implications, alter OpenAI’s tax status, and require approval from existing investors—including Microsoft, which holds a profit-sharing agreement and board observer rights.

Yet the mere fact that such talks are occurring illustrates the extraordinary nature of the AI moment. As artificial intelligence moves from laboratory experiment to foundational economic infrastructure, the rules of engagement between the state and the companies building that infrastructure are still being written. The OpenAI proposal is an attempt to put a down payment on one possible future—one in which government is a stakeholder, not just a watchdog.

Whether that future is desirable depends on one’s view of the state’s proper role. To optimists, it represents a pragmatic blending of capital and policy, accelerating innovation while ensuring the public shares in the rewards. To pessimists, it is a shortcut that erodes the very independence that legitimate regulation requires. The coming months will reveal whether the idea catches fire or fizzles, but it has already reshaped the debate over who should own the engines of the next industrial revolution.