In partnership with
In June 2026, Senator Bernie Sanders introduced a bill requiring the largest artificial intelligence companies to hand half their stock to a new federal fund. A few weeks later, the Financial Times reported that OpenAI had floated a much smaller version: each leading AI developer would give 5% of its equity to a public fund modeled on Alaska's oil savings account. A democratic socialist senator and the head of the most valuable AI startup now agree on the basic shape of an idea, even as they disagree sharply on its size.
That idea is an AI wealth fund: a government-held investment pool that owns shares in AI companies and pays the returns to the public. The case for it is straightforward. If AI raises output mainly by making machines and software more valuable, the gains will flow to whoever owns those machines and that software. A wealth fund spreads that ownership. Whether it works depends less on the principle than on the details: how the shares are acquired, who controls them, and whether AI profits turn out to be as large as today's prices assume.
AI's gains flow to whoever owns the capital.
Economists use the word Capital for the machines, buildings, and software that businesses use to produce goods and services. AI systems are capital in this sense. A trained model, the data centers that run it, and the chips inside them are long-lived assets that a company owns and earns money from.
New capital usually raises Productivity, which is how much output an economy gets from each hour of work. Higher productivity is the main reason living standards rise over time. But it does not decide who receives the extra income. That depends on whether a technology makes workers more valuable or replaces part of what they do.
Earlier technologies tended to lift wages along with profits, because workers who used the new tools became more productive and employers competed to hire them. Many economists worry AI could be different. If software takes over a growing share of tasks, more income could go to the owners of that software and less to workers. Economists track this split as the labor share of income, which in the United States has drifted lower for decades, well before modern AI.
Ownership of AI capital is also concentrated. A handful of companies build the leading models, and their shares belong mostly to founders, venture investors, large tech firms, and households that already own significant stock. About four in ten American families own no stock at all, directly or through retirement accounts. That gap is the problem an AI wealth fund is meant to address.
How a fund turns AI profits into a public dividend
The mechanics come down to four steps. First, the government acquires shares. It can buy them with tax revenue, accept them as a gift, or collect a tax paid in stock instead of cash. Sanders's American AI Sovereign Wealth Fund Act takes the third route: a one-time tax of 50% of the stock of companies earning at least $200 million a year from AI, with firms that also run non-AI businesses required to split the two. OpenAI's reported idea takes the second route, with voluntary 5% stakes.
Second, the fund holds the shares and collects what it earns through dividends or rising share prices. Sanders's bill bars the fund from selling its stock, so the public keeps its stake as the companies grow.
Third, the fund pays out a fixed fraction each year. The Sanders bill allows spending 5% of the stock's average market value annually, the same kind of rule Alaska uses, and basing the draw on a multi-year average smooths payouts when markets swing and protects the principal for future generations.
Fourth, the money reaches people, either as an equal cash dividend to every resident or as revenue for public programs. This is where the fund meets Fiscal Policy, meaning government decisions about taxing and spending. A cash dividend directly raises household incomes. Money routed into the budget can instead pay for schools or health care, or reduce the need for other taxes.
This makes a wealth fund an unusual form of Wealth Redistribution. Most redistribution moves income after it is earned, by taxing one group and paying benefits to another. A wealth fund moves ownership itself, so the public automatically receives a share of future profits, whatever their size turns out to be.
Want to Be the Investor Who Saw It Coming?
Not just someone who backed the next company. Someone who backed something the world actually needs.
More people need more protein. But producing it is getting harder and more expensive. And ultimately, consumers are paying the price.
DIT AgTech is tackling the problem at the source, helping ranchers produce more with less and making the livestock supply chain more efficient, from the ranch to the dinner table.
Because feeding the world isn’t just about producing more food. It’s about making it possible to produce it efficiently enough for people to afford it.
That’s a mission worth backing.
And if you’ve landed on this ad, maybe it’s not by accident.
Because what you’re seeing now could be obvious to everyone else later.
At $1 a share, this is your sign.
Grab the opportunity. Be among the first.
And one day, you can say:
“I saw it before they did.”
𝘐𝘯 𝘮𝘢𝘬𝘪𝘯𝘨 𝘢𝘯 𝘪𝘯𝘷𝘦𝘴𝘵𝘮𝘦𝘯𝘵 𝘥𝘦𝘤𝘪𝘴𝘪𝘰𝘯, 𝘪𝘯𝘷𝘦𝘴𝘵𝘰𝘳𝘴 𝘮𝘶𝘴𝘵 𝘳𝘦𝘭𝘺 𝘰𝘯 𝘵𝘩𝘦𝘪𝘳 𝘰𝘸𝘯 𝘦𝘹𝘢𝘮𝘪𝘯𝘢𝘵𝘪𝘰𝘯 𝘰𝘧 𝘵𝘩𝘦 𝘪𝘴𝘴𝘶𝘦𝘳 𝘢𝘯𝘥 𝘵𝘩𝘦 𝘵𝘦𝘳𝘮𝘴 𝘰𝘧 𝘵𝘩𝘦 𝘰𝘧𝘧𝘦𝘳𝘪𝘯𝘨, 𝘪𝘯𝘤𝘭𝘶𝘥𝘪𝘯𝘨 𝘵𝘩𝘦 𝘮𝘦𝘳𝘪𝘵𝘴 𝘢𝘯𝘥 𝘳𝘪𝘴𝘬𝘴 𝘪𝘯𝘷𝘰𝘭𝘷𝘦𝘥. 𝘋𝘐𝘛 𝘈𝘨𝘛𝘦𝘤𝘩 𝘩𝘢𝘴 𝘧𝘪𝘭𝘦𝘥 𝘢 𝘍𝘰𝘳𝘮 𝘊 𝘸𝘪𝘵𝘩 𝘵𝘩𝘦 𝘚𝘦𝘤𝘶𝘳𝘪𝘵𝘪𝘦𝘴 𝘢𝘯𝘥 𝘌𝘹𝘤𝘩𝘢𝘯𝘨𝘦 𝘊𝘰𝘮𝘮𝘪𝘴𝘴𝘪𝘰𝘯 𝘪𝘯 𝘤𝘰𝘯𝘯𝘦𝘤𝘵𝘪𝘰𝘯 𝘸𝘪𝘵𝘩 𝘪𝘵𝘴 𝘰𝘧𝘧𝘦𝘳𝘪𝘯𝘨, 𝘢 𝘤𝘰𝘱𝘺 𝘰𝘧 𝘸𝘩𝘪𝘤𝘩 𝘮𝘢𝘺 𝘣𝘦 𝘰𝘣𝘵𝘢𝘪𝘯𝘦𝘥 𝘩𝘦𝘳𝘦: https://bit.ly/4bzuWCi
Alaska and Norway show the model can last, and where it strains
The closest American precedent is the Alaska Permanent Fund. Voters created it in 1976 to save part of the state's oil wealth, and it has paid an annual dividend to residents since 1982. The fund now holds about $89 billion. Starting October 1, 2026, nearly 600,000 Alaskans will receive a $1,000 dividend plus a $200 energy relief payment.
Alaska also shows a fund's political weak spot. The dividend formula written into state law has not been followed since 2016. After oil revenue fell, lawmakers began using a growing share of fund earnings to pay for state services, and they now set the dividend in an annual budget fight. The 2022 payment topped $3,200 with energy relief; recent checks have been far smaller.
Norway offers a larger and more insulated example. Its oil fund, run by the central bank, holds roughly $2 trillion in stocks, bonds, and real estate, and a fiscal rule limits how much the government may spend from it each year. Unlike the AI proposals, it deliberately spreads its money across thousands of companies worldwide instead of concentrating in the industry that created the wealth.
The United States has recently moved closer to public ownership of private firms. In February 2025, President Trump signed an executive order directing officials to plan a national sovereign wealth fund, and in August 2025 the government took a stake of about 10% in Intel. Neither step created a dividend for citizens, but both made government equity stakes a mainstream topic in Washington.
The payoff rests on valuations, design, and restraint.
The first risk is valuation. Sanders estimates his fund would start with about $7 trillion, based on current AI company prices. A 5% annual draw on that sum is about $350 billion, or roughly $1,000 per American. But those prices assume AI firms will earn very large profits later, and many leading labs currently spend far more on chips and data centers than they take in. A fund concentrated in one sector rises and falls with it. In the week the White House talks became public, AI-linked stocks lost more than $1 trillion in market value.
The second issue is scale. OpenAI's private valuation of $852 billion in March 2026 makes a 5% stake worth about $42.6 billion. A 5% annual draw on that would be about $2 billion, or roughly $6 per American. Even with matching stakes from other labs, a voluntary fund would be more symbolic than income-changing unless it grew substantially.
The third risk is behavior. A 50% stock tax is a large levy on investors. Critics argue it would discourage investment in American AI, push new firms to incorporate abroad, or lead companies to reorganize so their activities fall outside the bill's definitions. Supporters reply that a one-time tax changes future incentives less than a recurring one, though investors may fear it could be repeated.
The fourth risk is governance. Sanders wants the fund to use its voting shares to block harmful corporate decisions; critics see that as political control of private firms. Even a passive fund invites pressure, including fears that a government owner would rescue a failing AI company. The Sanders bill bans bailouts from the fund, but such rules last only as long as future Congresses respect them.
Finally, the productivity gains themselves are uncertain. A National Bureau of Economic Research study published in February 2026 found that about nine in ten firms reported no effect of AI on productivity so far, even as executives expected gains ahead. If the payoff arrives slowly, the profits a fund would share will be smaller than hoped.
The signals that will show whether the idea takes hold
The first signal is the outcome of the White House's talks with OpenAI. Any announced stake will show whether the goal is a dividend for citizens or simply a government position in the industry. The second is Congress. Sanders's bill, S.4825, faces long odds, but the November 3 midterm elections will decide which party controls Congress in 2027, and rival bills or hearings would signal wider support.
The third signal is AI company stock listings. Several AI firms are expected to go public, and public share prices would give the first market-tested values for the assets a fund would hold. The fourth is economic data. The Bureau of Labor Statistics' quarterly productivity reports and its measures of labor's share of income will show whether AI is raising output and who is capturing the gains. Rising productivity alongside a falling labor share would strengthen the case for a fund; rising productivity alongside rising wages would weaken it.
An ownership answer to an ownership problem
An AI wealth fund rests on a simple observation: if AI's gains flow mainly to the owners of AI capital, the most direct way to share them is to share the ownership. That logic now draws support from Bernie Sanders to Sam Altman, and Alaska has shown for more than four decades that a public fund can pay citizens a dividend.
The harder questions are about design. A voluntary 5% stake would pay a few dollars a year per person. A 50% stake could pay around $1,000, but it would impose a large tax and concentrate public savings in one volatile industry. Either version depends on AI delivering the profits its valuations assume, and on future lawmakers leaving the fund intact. The principle is coherent; its value will be settled in the details.
