Trump Eyes US Government AI Stake for National Security

The White House is in active talks with leading AI companies—led by OpenAI's proposal to grant the U.S. government a 5% equity share worth roughly $42.6 billion—as Washington fuses AI policy news with a sweeping national security doctrine centered on tech sovereignty and domestic AI supremacy.

Lead

The Trump administration is pursuing a direct financial claim on America's frontier artificial intelligence industry, with OpenAI formally proposing to grant the federal government a 5% equity stake worth approximately $42.6 billion. The offer, first reported by the Financial Times on July 2, 2026, is part of broader discussions between OpenAI CEO Sam Altman, President Donald Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Scott Bessent. The proposal reflects Washington's intent to treat AI infrastructure as a national strategic asset — a posture reinforced weeks earlier by the June 5 signing of National Security Presidential Memorandum-11, which reshapes how the entire U.S. national security enterprise adopts and deploys AI.

What Happened

Altman has pitched a structure in which leading American AI companies collectively allot 5% of their equity into a sovereign-style vehicle modeled on the Alaska Permanent Fund — a state-level fund that invests oil revenues and distributes annual dividends to residents, valued at nearly $91.2 billion as of May 31, 2026. In OpenAI's version, the vehicle would be a Public Wealth Fund, enabling ordinary Americans to participate financially in the AI boom. Trump has publicly endorsed the concept as "very American," signaling White House alignment.

Talks remain at the conceptual stage. Whether other major AI companies — including Anthropic, Google DeepMind, xAI, and Meta Platforms — would participate under similar terms is unresolved. Any formal implementation is expected to require Congressional approval, adding legislative complexity to what is already an unprecedented governance experiment.

Strategic Context

The equity stake discussions are inseparable from Washington's broader US AI security architecture. NSPM-11 orders the Department of Defense and the Intelligence Community to rapidly onboard the most advanced AI models from multiple vendors, build next-generation high-security computing facilities, and close the capability gap between what is available commercially and what national security agencies can access. The directive explicitly rescinds Biden-era National Security Memorandum-25, removing a set of restrictions the Trump administration characterized as obstructive.

In a parallel step taken in May 2026, the Department of Defense signed deployment agreements with eight leading AI companies to embed their systems directly into classified military networks — marking the first time frontier commercial AI models have been formally integrated into warfighter infrastructure at scale.

NSPM-11 also mandates that contracts with AI vendors who seek to limit government use of their products be terminated, either for default or convenience. This clause represents a significant shift in the government-vendor relationship: the federal government is no longer a passive customer but an assertive principal with operational authority over how contracted AI systems are used.

AI Policy News: The Ownership Lever

The equity stake proposal represents a novel instrument in the AI policy news landscape — one that bypasses traditional regulatory frameworks in favor of ownership-based governance. Rather than imposing restrictions through legislation or rulemaking, the administration is seeking a seat at the board table through equity, early access agreements, and deployment mandates.

This approach has historical antecedents in energy policy, where the federal government has long held royalty stakes in natural resource extraction. Applying a similar model to artificial intelligence reflects a judgment within the administration that AI compute and frontier models constitute a resource of equivalent strategic importance to oil and gas.

The Commerce Department's recent decision to lift export controls on two of Anthropic's most advanced AI models illustrates the administration's transactional regulatory posture: companies that align with Washington's national security agenda gain operating latitude; those that resist face procurement consequences.

Tech Sovereignty Dimension

At the geopolitical level, the Trump administration's tech sovereignty push is explicitly framed as a counter to Chinese AI development. NSPM-11 prioritizes domestic AI procurement and positions U.S. military AI capability as a deterrent. The Department of Defense's agreements with commercial AI firms are designed to prevent a scenario in which the People's Liberation Army operates AI-enabled weapons systems at a pace or scale that outstrips U.S. military responsiveness.

The proposed Public Wealth Fund also serves a soft-power function: by making American citizens financial beneficiaries of frontier AI, the administration seeks to build a domestic political constituency for aggressive AI investment and deployment — an approach that mirrors how sovereign wealth funds in Gulf states have created public alignment with state-led industrial policy.

Export control policy remains a live variable. The administration has oscillated between tightening restrictions — particularly regarding high-end GPU exports to China — and easing them for allied nations and select commercial partners, making the US AI security posture difficult to characterize as uniformly restrictive or permissive.

Outlook

The Trump administration's dual-track approach — securing equity stakes in frontier AI companies while simultaneously mandating their deployment across classified national security networks — represents the most direct assertion of federal authority over AI development since the technology became commercially dominant. Whether the equity structure advances through Congress, and whether other AI companies accept terms similar to OpenAI's proposal, will determine how far Washington's ownership model extends. The broader trajectory is clear: the federal government intends to be a stakeholder, not merely a regulator, in the AI industry it increasingly views as foundational to American power.