The case for an AI sovereign wealth fund

As artificial intelligence generates vast new fortunes, the question of who will ultimately benefit from that wealth remains unsettled. According to UN Trade and Development, the global AI market is projected to expand from $189 billion in 2023 to $4.8 trillion by 2033. Much of that value, however, is expected to accrue to a relatively small group of companies and investors concentrated in a handful of countries.Nvidia alone was valued at more than $5 trillion in 2025, while Microsoft, Alphabet, Amazon, and Meta are collectively committing hundreds of billions of dollars to AI infrastructure.

The concentration of economic power has prompted a broader debate over whether AI should be viewed solely as a private technology or as an asset whose benefits should be shared more broadly. Governments are already becoming participants in the AI economy, with Gulf sovereign wealth funds including Abu Dhabi's Mubadala and MGX and Saudi Arabia's Public Investment Fund committing billions of dollars to chips, data centers, and leading AI companies. At the same time, some economists and policymakers have proposed the creation of AI sovereign wealth funds, under which governments would acquire stakes in AI companies and infrastructure and use the returns to support public services, pensions, or even make direct payments to citizens.

As states increasingly move from regulators to investors, the debate is expanding beyond how artificial intelligence should be governed to who should ultimately own the wealth it creates. At stake is not merely the future of a technology industry, but the ownership of what may become one of the defining sources of wealth of the twenty-first century.

Who owns the gains from AI?

Much of the public debate surrounding artificial intelligence has focused on regulation, safety, and competition. Yet some scholars argue that ownership deserves equal attention.

In a 2024 research paper, New York University professor Amy Whitaker argued that regulating AI and owning AI are not the same thing. Governments can impose rules on companies and address concerns over safety or market power, but those measures do not necessarily determine who captures the wealth generated by the technology. Ownership structures, she argues, ultimately determine who benefits from AI's economic gains.

Viewed through that lens, the debate extends beyond antitrust and oversight. Should AI be treated like most technologies, with rewards flowing primarily to shareholders and venture capital investors, or should societies seek ways for the public to share more directly in the wealth it creates?

From oil wealth to AI wealth

The idea that extraordinary sources of wealth should benefit the broader public is hardly new. Alaska established its Permanent Fund in 1976 to invest a portion of the state's oil revenues, while Norway created its Government Petroleum Fund, now known as the Government Pension Fund Global, in 1990 to preserve petroleum wealth for future generations.

Some researchers argue that artificial intelligence may warrant a similar approach. A 2025 policy paper published by Convergence Analysis, a technological policy think tank, proposed that governments acquire ownership stakes in AI companies and infrastructure, allowing societies to participate more directly in the wealth generated by AI rather than relying solely on taxation after value has already been concentrated.

The proposal does not imply nationalization. Proponents envision governments acting as long-term investors rather than operators, holding stakes in AI companies much as traditional sovereign wealth funds invest in public markets without directing the businesses themselves. In that sense, supporters see AI sovereign wealth funds not as a substitute for markets, but as a way to broaden participation in the gains they generate.

Governments as AI investors

Elements of this model are already visible, particularly in the Gulf. In June 2026, Abu Dhabi-backed MGX has invested alongside BlackRock, Microsoft, and Global Infrastructure Partners in a $30 billion AI infrastructure initiative, while Mubadala has expanded investments across semiconductors and advanced technologies. Saudi Arabia's Public Investment Fund has also accelerated spending aimed at turning the kingdom into a major AI hub.

These investments are motivated partly by economics and partly by national strategy. A September 2024 report from Vanderbilt University's Policy Accelerator argued that advanced AI capabilities carry growing national security implications, making access to computing power and frontier models matters of strategic importance rather than purely commercial concerns.

These developments point to a changing role for governments. Rather than acting solely as regulators, states are becoming investors and, potentially, owners of the infrastructure underpinning artificial intelligence.

A political debate enters the mainstream

Ideas once confined to academic circles are beginning to enter mainstream public debate.

In a June 2026 New York Times opinion essay, Senator Bernie Sanders argued that taxpayers deserve ownership stakes in firms whose growth relies heavily on publicly funded research. He proposed public ownership stakes in leading AI firms and compared the idea to the Alaska Permanent Fund, which distributes annual dividends financed by oil revenues.

The debate has also begun to cross ideological lines. Also in June, President Donald Trump said he believed AI companies would eventually agree to "give back" to the public, reflecting growing interest in mechanisms that would spread the benefits of AI beyond investors and technology executives.

Yet broadening ownership is only one proposed response to the concentration of AI wealth, and the idea remains contested. Critics caution that sovereign wealth funds could invite political interference, distort markets, or discourage innovation. Others argue that taxation and antitrust policies are better suited to addressing concentration.

Regardless of whether AI sovereign wealth funds ever materialize, their emergence reflects a broader shift in how policymakers think about technology and ownership. For decades, technology has largely been treated as a private enterprise whose rewards accrue to shareholders. Artificial intelligence may force governments and societies to reconsider that assumption. As AI becomes more deeply embedded in economies around the world, the central issue may no longer be whether it should be regulated. It may be who owns the wealth it creates.