The Public Has a Claim

The home-care worker may never receive shares in a leading AI company. The parent providing unpaid care may have no employer from whom to ask. A small business can work hard, adopt better tools and still send much of its software bill to owners elsewhere.

If ownership depends only on the employment bargain, these people may spend their lives close to the work and far from the assets.

That is why the claim must become public as well as personal. Every member of the community should have a real stake in productive wealth. Employment can add to that stake. It should not be the admission ticket.

This is the central demand of my proposal: universal capital ownership, built through personal accounts and a common pool of assets held for the public. The program must deliver income people can use, preserve wealth for later years and make its managers answerable to the owners they serve.

I have called this broader architecture an Ownership Democracy Act. That is the name of my proposal, not an enacted law. Here I want to make its practical choices clear enough that a reader can ask where the money enters, what they would own and who could change the terms.

We should pursue more capable AI, stronger safeguards and faster useful research. We should also build the ownership system now. Asking people to wait for their share until the shape of the economy is settled is one way to ensure that it settles without them.

Membership, written down

Begin with the person rather than the investment product.

The law creating the program would define who belongs, what rights membership carries and how a person can enforce them. I favor a broad base that includes citizens and settled residents, with a funded capital account for children rather than a system that notices them only when they enter paid work. Residence rules would need clear definitions and fair treatment of movement across borders. They should not become an annual test of whether a person’s life is deserving.

People with low earnings, unpaid responsibilities or no capacity for paid work would be included. The right would survive a lost job. Its existence would not depend on pleasing an employer, supporting the government or completing a prescribed course.

There would be two linked forms of ownership.

The personal account would hold assets allocated to the individual under published rules. Employer contributions could join public contributions. Earned balances would remain attached to the person, subject to the withdrawal and transfer terms explained in the previous chapter.

The public fund would hold a separate portfolio collectively. Membership would carry rights to the distributions defined by law, information about management and a voice through accountable institutions. An individual would not be able to sell a fraction of the national portfolio to settle a private bill. The common asset would remain available for future members as well as those alive today.

These are different rights. The statement should say so. A public fund’s entire value cannot also appear as if it were a second collection of freely withdrawable personal balances. Nor should the same contribution be counted once as new fund capital and again as an individual asset without showing the actual transfer between them.

A cash dividend can itself be a lawful entitlement. Its payment need not be charity, and an enforceable income right can be stronger than a decorative share with no practical protection. My reason for adding assets is that ownership gives the program a continuing source of returns and a claim on future growth. We need to examine both the asset and the rights attached to it.

The public should receive an intelligible annual statement. It would distinguish personal holdings, collective holdings and cash paid. It would show fees and explain how the next payment is calculated. A person missing from the register would have a route to correction, including help that does not require owning a computer.

Universality is a commitment that the system must carry out. It is not achieved by putting the word in the title of an app.

Put assets behind the promise

A dividend program needs revenue before it needs a beautiful payment screen.

I propose several routes into the fund, chosen for the source of value involved. When public money helps finance a commercial venture, the agreement can include an ownership stake, a right to acquire shares or a success-based payment. When a business receives valuable access to public land or infrastructure, the terms can include an appropriate public return. A carefully designed charge on extraordinary profits can contribute where firms capture returns well beyond those needed to support investment and risk.

These routes should be specified in advance. A company deciding whether to accept public support should know the financial terms. The government should be able to explain why the return is fair and whether the support is worth providing at all.

None of this creates free money. A public investment can fail. A company may offer less elsewhere in a contract because it must provide a future share. A badly priced agreement can leave the public carrying risk while surrendering the upside. The aim is competent bargaining on the public’s behalf, with the cost and expected return visible.

Some public spending should continue to produce open knowledge rather than a commercial claim. Basic research can be worth funding because others can build on it freely. A government should not place a private toll on every idea merely to fill a dividend account. The return can take the form of widely available knowledge, better services or lower costs. Where the bargain deliberately builds a private commercial asset, a financial public claim deserves consideration too.

The tax side requires equal care. Economic rent means a return above what is needed to keep resources in the activity. It is different from calling every successful business’s profit excessive. The boundary is difficult to measure. Any proposed tax base needs transparent allowances, administration that can enforce it and evidence about how firms respond.

My target is large, concentrated gains and publicly enabled windfalls. Ordinary adoption should remain attractive. A small repair shop buying an assistant should not pay a special penalty for becoming better at its work. A low-margin firm should not be treated as if it owns the platform on which it depends.

This is why I reject a general charge on every task performed by AI. Such a charge can make useful production more expensive without finding the wealth we mean to share. It can also reward firms for disguising the tool they use rather than creating a better product.

Even a well-targeted levy must be tested for who finally bears its cost. The business writing the check may respond through prices, wages, investment or profits. Officials should examine those effects after implementation and adjust a failing design. A tax described as a levy on owners is not proved to fall only on owners by its name.

Competition matters here. If the program depends on preserving high prices so that a few firms keep generating large rents, it has acquired a bad incentive. We should welcome lower prices and stronger competition even when they reduce one funding stream. The public fund should hold diversified assets and receive more than one kind of revenue, rather than becoming the financial defender of a monopoly.

A real bill, and different possible claims

Public AI ownership is already the subject of a concrete federal proposal in the United States. On June 18, 2026, Senator Bernie Sanders introduced S. 4825, the American A.I. Sovereign Wealth Fund Act, which was referred to the Senate Finance Committee. It would require covered AI companies to issue new equity so that Treasury held half the outstanding equity after remittance. A commission would manage the fund. Its distribution provisions authorize appropriations; they do not establish a guaranteed personal check of a stated amount.¹

I welcome the seriousness of putting an ownership claim into legislative text. The details are open to argument, including the size of the claim, its effect on investment and how citizens would exercise authority over the resulting fund. My Ownership Democracy Act is a separate proposal. The shared ambition of broad public ownership does not make the designs identical.

Existing public funds also show why precise language matters. Norway’s Government Pension Fund Global holds assets for the public under a chain of authority running from parliament and government to its investment managers. It is collective wealth, not a set of personal shares each Norwegian can sell.²

Alaska’s Permanent Fund uses a draw based on a multiyear average of fund value. The statutory percentage and the amount available constrain the draw, which remains subject to appropriation.³ That fund is an important example of public assets and rules for distributing value. It is not evidence that a new AI fund would receive the same revenue or offer the same individual rights.

We can learn from existing institutions without pretending that the name wealth fund supplies the whole design.

For the program I favor, the public’s financial claim would be stated in law rather than left to an annual announcement of generosity. Citizens would be able to check the calculation and challenge failures to apply it. Legislators would retain the power to amend the law through a visible process. No honest designer can promise that an ordinary statute will be beyond every future government.

Durability comes from clear rights, protected assets, competent administration and a public able to see what is being changed. The next chapter takes up the political work of winning and defending that settlement.

Make the payment fit the money

A large fund value does not tell us how large a sustainable household payment will be.

Suppose a fund holds assets valued at $100 million and proposes to distribute $4 million in a year. With 100,000 equal recipients, that is $40 each before administration and other deductions. The arithmetic does not tell us whether the fund can afford that distribution. It does show why an impressive asset total can produce a modest personal payment.

Some assets may also be difficult to sell. Their estimated value can rise without enough cash entering the fund to meet a promised payment. A dividend policy must distinguish asset valuations from spendable receipts.

I would use a published distribution formula with a smoothing period, so one unusually strong market year does not become a permanent spending promise. Part of new receipts would build assets. A separate reserve would help manage short-term swings in payments. Its size and use would be public, with limits on how long it could cover a gap.

A reserve buys time. It cannot turn a permanently inadequate revenue stream into a sufficient one. If income falls for long enough, the program must change contributions, payments or other spending. The choices should be explained early rather than concealed behind optimistic valuations.

The formula could also respond to economic change. A sustained weakening of labor’s share of income could trigger a review or a defined increase in public capital contributions. The measure would need a stable definition, a smoothing window and protections against treating every recession as evidence of the same structural shift. Funding the response would remain necessary; a trigger does not manufacture assets.

Where an automatic contribution increase is used, it should be capped and designed for the business cycle. The goal is to strengthen household ownership as the distribution of income changes, without demanding the largest new payment from fragile businesses at the worst moment.

This program also needs a plain account of its relationship with existing support. A modest dividend should not be used as an excuse to withdraw essential assistance whose value is much larger. The rules should show whether and how a payment affects taxes or other benefits. A family should not discover after receiving the money that it has unexpectedly lost more elsewhere.

The same discipline applies to inflation. What matters to the household is what the payment can buy. A rising nominal dividend is less impressive if rent, energy and other essential costs rise faster. Ownership policy should sit beside efforts to expand supply and keep markets open, rather than becoming a reason to stop caring about prices.

Start with payments the actual resources support. Let the claim grow as the assets and revenue grow. A smaller promise kept is a stronger foundation than a large figure repeatedly postponed.

Who acts for the owners

A public portfolio needs professional management. It also needs owners who can call that management to account.

The legislature would set the mandate, the permitted level of risk and the main distribution rules. Managers would make investment decisions within those limits. They should not be ordered to buy a failing company because it belongs to a political ally, or to keep a poor investment simply because admitting a loss would be embarrassing.

An independent board should include a defined route for public representation alongside the expertise needed to oversee the assets. Selection, terms of service and removal rules would be public. Conflicts would have to be disclosed and managed. Independent audits and a body empowered to investigate wrongdoing would examine what the fund actually does.

Where the fund holds voting shares, its managers exercise voting power in the companies it owns. They should publish how they use it. A citizen should be able to see the policy, compare it with the record and challenge departures through the oversight system. A large pool of public shares must not become private political power for the people hired to manage it.

Personal accounts need similar protection against quiet extraction. Fees, transfers and missing contributions should be easy to inspect. Complaints need a decision-maker with power to correct the problem. The person receiving the statement should not have to win an argument with several disconnected contractors to recover a small amount that is rightfully theirs.

The test is whether these duties can be enforced when the officials and managers are ordinary people with mixed incentives. A system designed only for unusually virtuous leaders is unfinished.

The claim should reach the household

For the home-care worker, this could mean an asset balance that grows even when the employer’s margin is thin. For the unpaid caregiver, it could mean ownership that does not vanish because valuable work produces no wage. For a young person, it could mean entering adulthood with a stake and an understanding of the rights attached to it.

The amount matters. The control matters. So does the knowledge that the claim survives a change in work or government.

I want people to benefit when the economy becomes more capable without first having to prove that their suffering has become an emergency. That requires building the financial rights into the growth itself.

AI can help create more value. A public ownership system can give everyone a lasting claim on a share of productive wealth. Neither task completes the other. We should pursue them together, with enough ambition to build and enough precision to make the benefits belong to people in fact.

Sources and notes