Innovation Accounts: A Proposal for Sharing the Upside of AI Growth with Americans

Many Americans feel left out of the abundance generated by artificial intelligence, despite living in a country that is home to most of the world’s leading AI companies.

The idea of Innovation Accounts has been proposed as a way for more Americans to benefit from their nation’s ingenuity. Modeled after Trump Accounts, which give newborns a stake in the stock market, Innovation Accounts would put seed money into a technology-stock index on behalf of adults, giving each of them a share in America’s technological engine.

OpenAI has reportedly suggested that governments take up to a 5% stake in AI companies. Both President Trump and progressive Democrats have floated public ownership stakes in these firms as well. However, the idea is not for governments to pick stocks but rather for individuals to own the upside directly.

The math behind Innovation Accounts is straightforward: $1,000 invested in the Nasdaq-100 a decade ago would be worth $5,727 more today, assuming 21% annual returns. Seed ‘innovation accounts’ with $5,000 and that’s $28,635 in gains over the same period.

There is precedent for this idea - it’s an extension of one already on the books. Trump Accounts, created under the One Big Beautiful Bill Act, give American children born from 2025 through 2028 a $1,000 government-funded stake in the stock market. Alaska has run its own version of this program for decades: its Permanent Fund invests a share of the state’s oil wealth and pays nearly every resident an annual dividend.

A private-sector asset manager would oversee investments under Innovation Accounts, with participants able to cash out gains at specific intervals - potentially every couple of years. This approach is designed to give more Americans access to the upside of AI-driven productivity growth without exposing them to excessive risk.

Currently, just 58% of American adults own stock, much of it held in retirement accounts. That means most people are exposed to AI’s downsides - rising electricity costs and potential job disruption - but have no claim on its benefits. Innovation Accounts would close this gap by giving Americans a stake in the technology index.

The source and use of funds for Innovation Accounts will likely be subject to great debate. Federal or state governments could seed the accounts, but philanthropy offers a faster and deficit-neutral start. The beneficiaries of AI - founders, early employees, investors, and foundations at labs like OpenAI and Anthropic - are a natural source of capital for this idea given their significant liquidity.

One of the hardest design questions is eligibility: who should be eligible to participate in Innovation Accounts? Opening $1,000 accounts for every American over 18 years old would cost roughly $270 billion. Limiting it to households with wealth under $500,000 - about 70% of Americans - cuts that to $190 billion.

The program could also be targeted rather than universal, triggered by sector-specific unemployment spikes, income level, or geography. For example, an ‘unemployment trigger’ could target those in highly impacted sectors: if unemployment in your sector runs 1.5x the average unemployment level for the past decade, you qualify.

The practical question is scope: how big should Innovation Accounts be? A $5 billion pilot could seed 1 million accounts at $5,000 each.

An extra $6,000 or $29,000 in gains over a decade won’t replace a lost job but gives people capital to retrain, start a business, or navigate whatever comes next on their own terms. Philanthropic pilots can start compounding those returns today without waiting for Congress.