The White House is weighing taking ownership stakes in leading artificial intelligence firms to funnel the industry's profits to the public, President Donald Trump said on Air Force One on 5 June. He likened the idea to Washington's 10% equity purchase in Intel last year and said he would host a meeting with AI executives as soon as next week; senior officials have held preliminary talks, according to digital outlet NOTUS, with options ranging from voluntary donations of shares to government-backed mechanisms that could see returns channelled to public purposes, including possible direct payments to households.

"We're talking about it, where the American people can benefit from the success of AI, the American people will like it better," President Trump told reporters, framing the proposal as a way to give households a direct share in the gains from AI.

The talks have both a commercial and a political dimension. CNBC reports that OpenAI chief executive Sam Altman has discussed the concept with the White House for more than a year, and that OpenAI set out a version of the idea in an April policy paper proposing a "Public Wealth Fund" to hold donated equity and invest in diversified, long-term assets. CNBC also reported that OpenAI is valued at more than $850 billion by private investors and is preparing for an initial public offering, although no formal investment terms have been announced.

What the White House is proposing

White House officials, according to NOTUS, are parsing a spectrum of models. At one end sits voluntary donations of shares by companies that want to demonstrate public-minded governance. At the other sits government-backed mechanisms that would take stakes and channel returns into public purposes, potentially including direct payments to households. How companies would cede shares, whether through donation, structured gifts, or purchase, and how those shares would be valued remain unsettled.

The proposal follows a broader shift in the administration's posture toward critical technology industries. Last year the government took a 10% stake in Intel, and it has acquired positions in other chip, rare earth and quantum-related firms, according to multiple reports. The move into ownership is therefore both precedent and policy signal: it suggests the administration is prepared to widen its role beyond regulation and procurement and into direct financial participation in technology firms.

Mastheads

Practical and political obstacles

Any plan to give the state an ownership role in firms it may later regulate raises obvious conflicts. Critics warn that such holdings could create perverse incentives, concentrating influence over information and decision-making in the nexus of government and big tech. David Sacks, a former White House technology adviser, criticised the plan as accelerating "the corporate-government fusion we’re already sliding toward," arguing it risks expanding centralised influence over information and decision-making.

Security concerns are a second vector pushing officials toward greater involvement. Reuters reported heightened fears after Anthropic released a powerful tool called Mythos, with experts warning that such capabilities could enable sophisticated cyberattacks against sectors like banking that rely on complex legacy systems. Reuters and other outlets have said Anthropic, OpenAI, Google, Meta, Microsoft and SpaceX were among the companies mentioned in reporting on White House engagement, though many firms declined to comment publicly.

Legislative pressure is also a factor. Senator Bernie Sanders has proposed a bill that would give the government 50% ownership in the largest AI companies, a far larger stake than the administration has publicly discussed. The Sanders proposal adds political momentum on the left for big structural interventions, and it sharpens the debate about what scale of public ownership would be legally and politically doable.

The White House has already taken regulatory action on AI this term. Mr Trump last month rescinded an earlier planned executive order and then signed a redesigned directive asking leading AI developers to voluntarily submit their most capable models for government cybersecurity tests before public release. Administration officials say they're open to exploring concepts that would let Americans share in AI's economic upside, but they haven't released a valuation framework, legal pathway or detailed policy mechanics for acquiring equity.

On the corporate side, the idea isn't entirely new. CNBC reported that OpenAI discussed a public-wealth model as part of its outreach to policymakers, and the company's April paper envisaged a vehicle to hold donated equity and invest it for broad public benefit. But translating concept to contract will require agreement on valuation, governance and exit rules, and it will demand legal work to avoid undermining competition policy or creating regulatory capture.

Related Articles

Mr Trump is due to meet AI executives as soon as next week to test whether firms will entertain voluntary donations or resist a more formal government equity role, and to discuss how any returns might be channelled to the public. Originally reported by Reuters.

This article was created with AI assistance.