Trump Eyes US Stake in AI Companies: 9.9% Model [2026]

The Trump administration is weighing one of the most consequential interventions in the history of American technology policy: a government stake in AI companies that would give the U.S. Treasury – and, the White House argues, the American public – a direct financial claim on the firms building frontier artificial intelligence. President Donald Trump confirmed in early June 2026 that he expected to meet with “all the big ones” at the White House, and that the administration was exploring equity arrangements designed so that ordinary Americans “can benefit from the success of AI.” The disclosure landed in the middle of a sharp Big Tech selloff and reignited a fierce debate about whether Washington is drifting toward state capitalism.

1. What Trump Actually Proposed on the Government Stake in AI Companies
2. The Intel Precedent: A 9.9% Stake and $8.9 Billion
3. U.S. Government Equity Stakes in 2025–2026: The Track Record
4. The Big Tech Selloff: Market Reaction to the AI Stake Talk
5. Which AI Companies Are in the Frame
6. Competing Models: Sanders’ 50% vs. Trump’s Profit-Sharing vs. Sovereign Funds
7. The Capitalism vs. State Capitalism Debate
8. The White House AI Executive Order Backdrop
9. Historical Context: From Bailouts to Permanent Stakes
10. Market Impact and What’s at Stake for Investors
11. Predictions: Where the Government Stake in AI Companies Lands
12. Frequently Asked Questions
1. Is the U.S. government actually buying stakes in AI companies?
2. What is the precedent for a government stake in AI companies?
3. How big a stake is being discussed?
4. Why did tech stocks sell off?
5. Is a government stake in AI companies capitalism or socialism?
6. Do foreign governments already own parts of these companies?
7. When will we know the details?
8. Related Coverage
9. Sources & Further Reading

This is not an abstract trial balloon. The proposal sits on top of a concrete 2025 precedent – the federal government’s 9.9% stake in Intel, an $8.9 billion position converted from CHIPS Act grants – and a Pentagon equity deal with rare-earth miner MP Materials. What was once unthinkable for a free-market Republican administration has become a live policy menu. This analysis breaks down what is actually being discussed, the numbers behind the precedents, the market reaction, the competing models, expert reactions on both sides, and where a US government stake in AI companies could realistically land.

What Trump Actually Proposed on the Government Stake in AI Companies

According to reporting by Politico on June 5, 2026, the administration is exploring ways to give Americans an equity stake in AI developers to offset what officials describe as potentially “historic” labor-market disruption from automation. Trump told reporters he expected to host the heads of the largest AI companies at the White House the following week, signaling that the conversation had moved from think-tank whiteboards to the Oval Office. The framing is populist: if artificial intelligence generates enormous wealth and simultaneously eliminates jobs, the government wants a mechanism so the public shares in the upside rather than absorbing only the downside.

The specifics remain deliberately fluid. Reporting indicates the ideas range from direct equity ownership to profit-sharing arrangements and sovereign-fund-style holdings. Notably, the concept was reportedly floated by OpenAI itself, and discussed more broadly with Anthropic and Elon Musk’s xAI – meaning at least some of the companies in frame are not uniformly hostile to the idea. CNBC reported the administration was specifically exploring a position in OpenAI, with “possibly other AI giants as well.” No percentage, dollar figure, or binding term sheet has been published, and the White House has not committed to any single structure.

That ambiguity matters. A 1% sovereign-fund position is a fundamentally different policy than the 50% public ownership floated on Capitol Hill, and the gap between those poles defines the entire debate over a government stake in AI companies.

The Intel Precedent: A 9.9% Stake and $8.9 Billion

Every conversation about an AI equity stake begins with Intel. On August 22, 2025, Intel and the Trump administration announced a “historic agreement” under which the U.S. government acquired a 9.9% stake in the chipmaker. Per Intel’s own disclosure and its SEC filing, the government committed roughly $8.87 billion – $8,869,800,000 to be exact – in an investment in Intel common stock. Intel characterized the deal as part of an $11.1 billion total relationship once the $2.2 billion in CHIPS Act grants already received was included. Crucially, the cash was not fresh appropriation; it was largely a conversion of previously awarded grant money into equity.

Trump has explicitly cited the Intel arrangement as the template for the AI proposal, pointing to the existing roughly 10% position as proof the model works. The structure is instructive: the government took non-voting common stock plus a warrant, a deliberately passive posture intended to capture financial upside without day-to-day operational control. Supporters argue this is the cleanest way to give taxpayers a return on the billions in subsidies, tax incentives, and infrastructure that underpin the AI buildout. Critics counter that even a “passive” 10% federal holder distorts incentives, invites political pressure on hiring and pricing, and blurs the line between regulator and shareholder. The Intel deal is the proof of concept – and the warning sign – that the AI debate now orbits.

U.S. Government Equity Stakes in 2025–2026: The Track Record

Intel was not an isolated experiment. Over 2025, the administration assembled a pattern of direct government equity positions across strategically sensitive industries, each with a distinct structure. The table below summarizes the confirmed deals and the proposed AI extension.

Sources: Intel SEC filing (Aug 2025); C&EN, FAS (Jul 2025); public reporting. Figures reflect 2025–2026 disclosures.

The MP Materials deal is the most aggressive of the confirmed cases. In July 2025, the Department of Defense agreed to pay $400 million for an effective 15% stake in the rare-earths miner via Series A preferred stock and a warrant, making the U.S. government the company’s largest shareholder. The Pentagon paired the equity with a price-floor guarantee – protection against China flooding the market – and the Office of Strategic Capital added a separate $150 million loan. The U.S. Steel arrangement, tied to the Nippon Steel acquisition, took a different form: a “golden share” granting Washington veto rights over key decisions rather than a cash equity purchase. Together, the three deals show an administration comfortable using equity, preferred stock, and governance rights interchangeably depending on the strategic goal.

The Big Tech Selloff: Market Reaction to the AI Stake Talk

The timing was combustible. The equity-stake disclosure coincided with a broad technology selloff in the second week of June 2026, with Bloomberg framing the AI-ownership question as part of a “broad tech rout” that pushed some Asian indexes toward bear-market territory. Markets dislike ambiguity, and the prospect of the federal government as a shareholder – potentially diluting existing investors or attaching strings to operations – injected precisely the kind of policy uncertainty that compresses valuations on the richest-priced names in the market.

The sensitivity is structural. AI infrastructure names had carried the entire market for two years: Nvidia became the first company in history to cross a $4 trillion market capitalization in 2025, and the so-called Magnificent Seven accounted for an outsized share of S&P 500 gains. When a single policy headline can plausibly reprice the terms on which these companies raise capital and deploy it, the index-level fragility becomes obvious. Investors are not necessarily betting the proposal becomes law; they are pricing the tail risk that government ownership reshapes the return profile of the most valuable cohort of stocks on earth. For a sector trading at premium multiples on the assumption of light-touch regulation, even an exploratory equity stake is a material change in the risk calculus.

Which AI Companies Are in the Frame

The list of companies reportedly in scope reads like a who’s-who of the AI economy. OpenAI – which, per reporting, floated the profit-sharing concept – and Anthropic are the headline names, both described as on track for IPOs that reporting suggested could each be valued at over $1 trillion. Musk’s xAI was part of the broader discussion, and CNBC’s framing left the door open to additional “AI giants.” The valuations involved are staggering and help explain why a percentage point of equity is worth fighting over.

Sources: Politico (Jun 2026), CNBC, prior tech-insider reporting. Valuations are reported figures and vary by transaction type.

One wrinkle complicates the “America First” framing: some foreign governments are already shareholders. Per CNBC, several sovereign wealth funds – including Gulf-state vehicles – have invested in OpenAI and Anthropic. That reality cuts two ways in the policy debate. Proponents ask why foreign states should own slices of America’s most strategic companies while the U.S. Treasury owns none. Opponents warn that matching foreign sovereign behavior is a race to the bottom that politicizes capital allocation and undermines the very market dynamism that produced the AI boom in the first place.

Competing Models: Sanders’ 50% vs. Trump’s Profit-Sharing vs. Sovereign Funds

There is no single “government stake” proposal – there are at least four distinct models on the table, separated by an order of magnitude in ambition. Senator Bernie Sanders has pitched legislation that would create a 50% government ownership stake in some AI companies, an explicitly redistributive framework. Trump’s profit-sharing concept is far more modest and modeled on the passive Intel position. Then there are the foreign sovereign-fund holdings already in place, and the golden-share governance model used at U.S. Steel. The table maps the spectrum.

Sources: Politico, CNBC, public reporting (2025–2026).

The political distance between these models is enormous. A 9.9% passive Intel-style stake can plausibly be sold as capitalism with a national-interest twist. A 50% controlling position is, by any conventional definition, partial nationalization of the most dynamic sector of the U.S. economy. The administration will almost certainly gravitate toward the low end of the spectrum – but the existence of the Sanders proposal gives the entire conversation a gravitational pull leftward that would have been inconceivable in a Republican administration even two years ago.

The Capitalism vs. State Capitalism Debate

The fiercest fight is semantic and ideological: is a federal equity stake capitalism, or socialism by another name? Commerce Secretary Howard Lutnick has been the administration’s loudest defender of the model. Reacting to the Intel deal, Lutnick told CBS News bluntly: “Intel agreed to give us 10% of their company … So, it’s not socialism. This is capitalism.” The argument is that the government is acting like any other investor – taking equity in exchange for capital and support, capturing upside on behalf of taxpayers who funded the subsidies in the first place.

Free-market economists are unconvinced. Adam Posen, president of the Peterson Institute for International Economics, characterized the Intel move as “a step toward state capitalism,” warning that government ownership opens the door to political pressure and operational inefficiency. Daniel Di Martino, a fellow at the Manhattan Institute, was more pointed, telling CBS News the stake “certainly gets us closer [to socialism] and makes us less prosperous,” and that “we are intervening in the capital markets in a way that is going to lead to inefficiencies.” The critique is not merely ideological: a government shareholder that simultaneously wields regulatory and antitrust authority over the same company faces conflicts no private investor does. When the referee also owns the team, the integrity of the market is at stake.

The White House AI Executive Order Backdrop

The equity proposal does not exist in a vacuum. On June 2, 2026, the White House published a presidential action titled “Promoting Advanced Artificial Intelligence Innovation and Security,” which laid out the administration’s broader posture toward frontier models. The order asserts that the administration has “unleashed tremendous technological growth and economic investment in AI by slashing the bureaucratic constraints” of the prior era – a deregulatory framing that sits in tension with the interventionist equity idea.

Two provisions stand out. The order seeks federal access to covered frontier models up to 30 days before they are released to other trusted partners, an unusual early-access arrangement that effectively positions the government as a privileged tester of the most powerful systems. At the same time, it explicitly states that nothing in the relevant section authorizes “a mandatory licensing, preclearance, or permitting requirement” for AI models – a clear signal that the administration wants leverage and insight without the heavy regulatory apparatus critics fear. Read alongside the equity discussion, the order reveals a coherent if unconventional strategy: deregulate the operating environment, but secure the government a seat at the table through ownership and early access rather than through traditional rulemaking.

Historical Context: From Bailouts to Permanent Stakes

Government equity in private companies is not new in American history, but the context here is. During the 2008–2009 financial crisis, the Treasury took stakes in banks and automakers through TARP – but those were emergency, time-limited rescues explicitly designed to be unwound, and they generally were. The 2025–2026 deals are different in kind: they target healthy, often dominant companies, and they are framed as strategic and durable rather than as temporary crisis interventions. CNBC noted the administration had taken or explored direct positions across a range of firms including Intel, IBM, GlobalFoundries, and quantum-computing companies, suggesting a systematic approach rather than ad hoc deals.

This shift – from rescuing failing firms to co-owning winning ones – is the genuinely novel development. It echoes the sovereign-wealth-fund model pioneered by Norway, Singapore, and the Gulf states, where governments hold permanent diversified equity portfolios. Whether the U.S. is consciously building toward an American sovereign wealth fund or simply improvising deal-by-deal remains the central unanswered question. The AI proposal, if it advances, would mark the moment the experiment moved from strategic minerals and legacy chipmakers to the crown jewels of the U.S. equity market.

Market Impact and What’s at Stake for Investors

For investors, the proposal introduces a new and hard-to-hedge variable. The immediate concern is dilution and control: a primary-share issuance to the government, as happened with Intel, dilutes existing holders, while preferred structures or warrants can sit senior to common equity. The deeper concern is the precedent itself. If a government stake in AI companies becomes normalized, the market must reprice the assumption that America’s most valuable companies operate free of direct state ownership – an assumption baked into U.S. equity premiums for generations.

There is also a competitive dimension. A government stake could come bundled with benefits – preferential federal contracts, regulatory forbearance, or subsidized capital – that advantage the chosen firms over rivals. That prospect worries antitrust observers, who note that the AI market is already concentrated among a handful of well-capitalized players. The hyperscalers backing these labs are pouring well over half a trillion dollars combined into AI infrastructure in 2026 by most analyst estimates, and folding government ownership into that already-concentrated structure could entrench incumbents further. The investor takeaway is not panic but vigilance: the rules of engagement for owning AI exposure are being rewritten in real time.

Predictions: Where the Government Stake in AI Companies Lands

Based on the trajectory of the 2025 precedents and the political dynamics, several outcomes look most probable over the next 12 to 18 months:

• A passive, Intel-style minority stake is the most likely outcome. Expect any AI equity deal to land near the 5–10% range with non-voting structures, not the 50% Sanders model. The administration will prioritize the optics of “taxpayers benefit” over operational control.
• OpenAI is the most probable first mover. Because the concept was reportedly floated by OpenAI and the company is pursuing a public listing, it is the natural test case for any voluntary equity arrangement.
• Legal and constitutional challenges are near-certain. Any compelled or quasi-compelled federal equity stake will draw litigation over takings, due process, and the limits of executive authority absent fresh congressional appropriation.
• Market volatility persists as a policy-risk premium. Until concrete terms emerge, expect AI and semiconductor names to trade with an added headline-risk discount whenever equity-stake reporting resurfaces.
• The sovereign-wealth-fund framing gains traction. Expect the administration to increasingly package these deals under a coherent “American sovereign fund” narrative rather than defending them one company at a time.

Frequently Asked Questions

Is the U.S. government actually buying stakes in AI companies?

As of June 12, 2026, no deal has been finalized. The administration is exploring the idea, and Trump confirmed he expected to meet with major AI company leaders at the White House. The concept reportedly originated in part with OpenAI. No percentage or dollar figure has been published, and nothing is binding yet.

What is the precedent for a government stake in AI companies?

The clearest precedent is Intel. In August 2025 the U.S. government acquired a 9.9% stake worth roughly $8.9 billion, largely by converting CHIPS Act grants into equity. The Pentagon also took an effective 15% stake in rare-earth miner MP Materials for $400 million in July 2025. Trump has cited the Intel deal as the model for any AI arrangement.

How big a stake is being discussed?

It depends on the proposal. Senator Bernie Sanders has pitched legislation creating up to a 50% government ownership stake in some AI firms. The administration’s own profit-sharing concept appears far more modest and modeled on the passive Intel position. No official figure has been set for the AI proposal.

Why did tech stocks sell off?

The equity-stake disclosure coincided with a broad tech rout in early-to-mid June 2026. Investors dislike the policy uncertainty around potential dilution, government control, and strings attached to operations – especially for premium-valued AI and chip names that had led the market for two years.

Is a government stake in AI companies capitalism or socialism?

That is the core debate. Commerce Secretary Howard Lutnick argues it is capitalism – the government investing for a return. Critics including the Peterson Institute’s Adam Posen call it “a step toward state capitalism,” warning of political pressure and market inefficiency. The answer hinges heavily on how large and how controlling any eventual stake turns out to be.

Do foreign governments already own parts of these companies?

Yes. Per CNBC reporting, several sovereign wealth funds – including Gulf-state vehicles – already hold positions in OpenAI and Anthropic. Proponents use this to argue the U.S. Treasury should also participate; critics say matching foreign sovereign behavior politicizes capital allocation.

When will we know the details?

Potentially soon. With White House meetings reportedly scheduled and an active executive-order framework already published in June 2026, concrete terms could emerge over the following weeks or months – though any binding deal will likely face legal challenges and congressional scrutiny first.

Related Coverage

Sources & Further Reading

This article is for informational purposes only and does not constitute investment advice. Figures cited reflect reporting and disclosures from 2025–2026 and are subject to change. Last updated June 12, 2026.

Nadia Dubois

AI & Innovation Editor

Nadia Dubois is the AI & Innovation Editor at Tech Insider, where she tracks the rapid evolution of artificial intelligence, from foundation models to real-world enterprise deployment. She previously covered AI and startups for La Tribune and contributed to MIT Technology Review’s European coverage. Nadia specializes in generative AI, AI regulation, and the intersection of technology and European industrial policy. She holds a dual degree in Computational Linguistics and Journalism from Sciences Po Paris.

View all articles



Content Curated Originally From Here