Technology stocks and the artificial intelligence industry have expanded rapidly, but many U.S. households do not own the assets benefiting from that growth. Technology investor and economist Sarah Cannon has proposed “Innovation Accounts,” which would give adults seed capital to invest in technology stock indexes. The aim is to broaden household exposure to AI-driven productivity and wealth gains while providing a financial buffer if employment is disrupted.
Tech returns are rising, but ownership remains limited
In the first half of 2026, hedge funds focused on technology delivered an average return of 27%, while the Nasdaq-100 gained about 20% over the same period. Cannon argues that AI-related growth and wealth creation are accelerating, but the gains are flowing mainly to people who already own technology assets. If AI adoption continues to expand, households without market exposure may find it increasingly difficult to share in those returns.
The ownership gap is also tied to potential labor-market adjustments. Households could face higher electricity costs or pressure on employment without receiving corresponding gains from asset prices. An April Gallup survey found that 58% of U.S. adults owned stocks, with a significant share of those holdings kept in retirement accounts.
A personal index account, not government stock picking
Cannon’s proposal would invest the initial capital in a technology stock index, with private asset managers handling the investments under government oversight. Account holders could withdraw gains on a set schedule, such as once every few years. She argues that this structure would be better suited to expanding personal participation than having the government directly own shares in AI companies, because the government is not well positioned to select individual stocks.
The proposal comes as OpenAI has suggested that the government could hold as much as a 5% stake in the company and that other leading AI firms could adopt a similar arrangement. President Donald Trump and progressive Democrats have also discussed giving the public ownership interests in AI companies. Cannon’s approach instead emphasizes individually held accounts and diversified index investments, rather than making the government a direct corporate shareholder.
As an illustration, Cannon said that a $1,000 investment in the Nasdaq-100 made 10 years ago would have gained $5,727 based on an annualized return of 21% over the past decade. A $5,000 initial investment would have produced gains of about $28,635. The examples are based on historical performance and do not guarantee future returns. Technology indexes can perform differently over time, and actual account balances would fluctuate with the market.
Trump Accounts and Alaska offer precedents
The concept builds on the children’s investment-account model enacted under the One Big Beautiful Bill Act. The “Trump Accounts” will provide $1,000 in government-funded stock-market investment capital for U.S. children born from 2025 through 2028. Cannon wants a similar mechanism extended to adults, giving them an asset that can accumulate as AI changes the economy.
Alaska’s Permanent Fund provides another reference point. The state invests part of its oil wealth and distributes annual dividends from the returns to nearly all residents, including children. The funding for Innovation Accounts has not been determined. Federal or state governments could provide the initial capital, while Cannon says philanthropic donations could launch pilot programs more quickly without directly increasing budget deficits. Founders, early employees and investors in the AI industry, along with foundations linked to laboratories such as OpenAI and Anthropic, could be potential contributors.
Universal coverage would carry a high cost
The scope of eligibility is a central design question. Providing $1,000 to every U.S. adult aged 18 or older would cost an estimated $270 billion. Limiting eligibility to households with less than $500,000 in wealth, covering roughly 70% of Americans, would reduce the cost to about $190 billion.
Another option would target payments by income, location or industry-specific unemployment. Cannon offered a “job-loss trigger” as one example: workers in an industry would qualify if unemployment measured under its NAICS industry code reached 1.5 times that industry’s average level over the previous 10 years. She also suggested that economists could develop more precise eligibility criteria.
The broader idea of universal basic capital has drawn support from figures with differing political positions, including MIT economist David Autor, OpenAI Chief Executive Sam Altman and Senator Bernie Sanders. The size of any program would remain a critical issue. A $5 billion pilot, after administrative costs, could provide $5,000 in seed capital to 1 million people. If the pilot proved effective, government or philanthropic funding could expand it gradually.
The accounts would supplement wages, not replace them
Cannon acknowledges that $6,000 or $29,000 in investment gains after 10 years would not replace the income lost from a job. She sees the funds instead as capital for retraining, starting a business or navigating a period of employment transition, giving affected workers more choices. A philanthropic pilot could begin without waiting for Congress to pass legislation.
The Nasdaq-100 and other technology-heavy indexes are not guaranteed to deliver a 21% annualized return over the next decade. The eventual scale of AI-related job losses is also uncertain. Cannon’s argument is that a funding mechanism can be established before those outcomes are clear. If the benefits of AI remain concentrated among a small group of asset owners, disputes over how the gains from technological change are distributed could intensify. The proposed accounts would give more households a financial stake in those gains while they absorb the effects of the transition.