Vice President JD Vance disclosed that the Trump administration is evaluating a sovereign wealth fund structure designed to provide US citizens direct equity stakes in artificial intelligence companies, marking the first time a sitting administration has proposed federal ownership positions in private AI assets. The remarks came as Elon Musk publicly advocated for direct cash payments to Americans from AI-related government revenues, revealing a split within the administration on wealth distribution mechanisms.
The sovereign fund proposal would establish a federal vehicle to acquire and hold minority stakes in AI firms, with ownership shares distributed to American citizens through a mechanism not yet defined. Vance offered no specifics on funding sources, governance structure, or which AI companies might be targeted, but the statement alone shifts the conversation from regulatory oversight to direct capital participation. The administration has not disclosed whether this would involve new equity purchases, negotiated stakes in exchange for regulatory concessions, or mandatory equity transfers from firms accessing federal compute resources. No draft legislation has surfaced, and Treasury has issued no formal comment.
The addressable market is considerable. US-headquartered AI companies raised $97B in venture funding in 2024, with valuations for OpenAI, Anthropic, and xAI alone exceeding $250B in aggregate. A sovereign fund acquiring even 2% stakes across the top fifteen privately held AI firms would require $10B to $15B in deployment capital, assuming pre-IPO valuations hold. The United States currently operates no true sovereign wealth fund; the closest analog is the Alaska Permanent Fund, which holds $78B and distributes annual dividends to residents from oil revenues. Scaling that model to AI equity at the federal level introduces questions of valuation, liquidation rights, and whether stakes would be transferable or held in perpetuity.
Musk's counter-proposal for direct cash payments reflects a cleaner but less durable approach. One-time distributions avoid the governance complexity of fund management but provide no ongoing exposure to AI upside. The philosophical divide is meaningful: Vance's model treats AI as a national resource requiring long-term equity participation, while Musk's treats it as a windfall requiring immediate redistribution. Both assume the federal government will capture substantial revenues or equity from AI development, an assumption not yet reflected in current policy. No tax structure, royalty framework, or compute-access levy has been codified.
Allocators should monitor three channels over the next 90 days. First, whether Treasury or the National Economic Council releases a formal feasibility study on fund structure and capitalization. Second, whether any AI firms signal willingness to negotiate equity stakes in exchange for regulatory clarity or access to federal compute infrastructure. Third, whether Senate Finance or House Ways and Means schedules hearings on AI revenue capture mechanisms, which would indicate legislative intent beyond trial balloons. The Alaska Permanent Fund took four years from proposal to first distribution; a federal AI fund would face longer implementation timelines unless structured as an executive vehicle bypassing appropriations.
The proposal arrives as global sovereign funds already hold $12.4T in assets, with Norway's Government Pension Fund and Abu Dhabi's ADIA deploying into late-stage AI rounds. A US federal fund would enter as a price-insensitive buyer with indefinite hold periods, distorting private valuations and exit timelines for existing investors.