Sam Altman and Elon Musk want to design your safety net. It's complicated.
Tech billionaires have a lot of ideas about how to give you money in the AI economy.
TL;DR: While tech leaders propose “compute credits” or “universal high income” as AI safety nets, private models risk replacing democratic citizenship with corporate dependency. True redistribution requires public legislation and progressive taxation.
“I no longer believe in universal basic income as much as I once did,” Sam Altman said last week, walking back his most famous policy idea.
A fixed cash payment, in his words, “does not get at what we’re really going to need.” His new proposal is “collective ownership that could be in compute or in equities or something else.” In other words: instead of the government giving you money, OpenAI proposes citizens receive stakes in AI-driven assets.
An OpenAI policy paper released in April proposes a public wealth fund that gives citizens a stake in AI-driven growth.
Two weeks earlier, Elon Musk started posting in all caps about what he calls “Universal HIGH INCOME”: not a basic safety net, but a high standard of living for everyone, funded by AI productivity gains so enormous that scarcity itself disappears.
Two of the world’s wealthiest men, two different proposals, both presented as solutions to a problem their companies are accelerating. And designed by them, for you, with the redistribution mechanism running through infrastructure they control.
But there’s already a perfectly good way to redistribute the wealth that AI is generating. It’s called taxation.
What is UBI, anyway?
The term UBI gets used loosely, but the cleanest definition, used by the World Bank, is that it’s a regular cash payment given to individuals with few or no conditions.
UBI is easy to explain but difficult to implement. Its design questions are political, not technical. Who pays? Who gets it? These are decisions that legislatures make. They’re the kinds of decisions that get harder when the people steering the conversation are the ones who'd owe the most under any honest version of it.
In Canada the live policy conversation is closer to a guaranteed basic income, which targets support to people below a certain income level and phases out as income rises. Bill S-206, currently before the Senate, is framed around developing a national framework for guaranteed livable basic income.
From UBI to AI productivity gains
UBI isn’t new. US Founding Father Thomas Paine sketched a version in 1797 and Dr. Martin Luther King Jr. argued for guaranteed income. For most of history, basic income was a demand made of the powerful, not an offer made by them.
In the last decade, though, the explosion of generative AI rekindled the conversation. If AI makes the economy more productive, who shares in those gains? Right now, it’s shareholders and infrastructure owners.
As the public’s approval of AI continues to drop, the tech industry is attempting to mitigate the backlash. But their proposals share a common feature: the redistribution flows through infrastructure they own. The proposals are being floated as alternatives to taxing tech firms.
A lone AI CEO argues for taxation
The strongest in-industry case for actual taxation comes from Anthropic CEO Dario Amodei. Speaking to Axios in January, Amodei said: “We have enormous wealth, but distribution is a problem.”
He argued that AI will create trillionaires, that current tax frameworks weren’t built for that scale of concentration, and that taxation is a stability question, not just an economic one. Done wrong, he warned, “You’re going to get a mob coming for you.”
Amodei doesn’t offer any specific models. But an AI CEO arguing on the record that the answer has to be taxation is the kind of argument that much of the tech industry avoids.
What basic income studies find
Setting the political question aside, the case for unconditional cash, on its own, is reasonably solid.
Stockton’s Economic Empowerment Demonstration gave $500 a month to 125 residents over two years. In the first year, full-time employment among recipients rose at more than twice the rate of the control group. Anxiety and depression dropped. Recipients used the cushion for car repairs and childcare, providing access to better jobs.
Finland’s two-year pilot for 2,000 unemployed people produced no significant change in employment but clear gains in life satisfaction, and reduced stress.
The OpenResearch study (funded by Altman) gave 1,000 low-income participants $1,000 a month for three years. Recipients worked about 1.3 fewer hours a week, increased spending on healthcare and education, and showed a brief well-being spike that faded by year 2.
But the findings don’t tell us what happens at a national scale, indefinitely, with the labour market fundamentally restructured. Pilots are evidence, not a plan. Going from “a pilot showed promise” to “this is the post-AI social contract” is difficult.
Public citizenship vs. private subscription
Compare what democratically administered cash gives you with what a compute allowance delivers.
Cash you can spend on anything: Rent, groceries, donations to a campaign against the people who paid the taxes that funded it. Cash is a resource you can use to make the system better, including voting for politicians who want to fortify it.
A compute stake from OpenAI: Benefits tied to AI assets, distributed publicly but originating from company-influenced policy. You can’t pay rent with compute. It’s an asset right, not unrestricted cash.
A government cheque is administered by an elected legislature, paid for by taxes you can debate, set, and revise through political action. If you don’t like a program, you vote for someone who’ll change it. If it’s underfunded, you can organize for higher taxes.
This is the difference between being a citizen and being a customer.
What a public version looks like
A public version looks like Norway’s sovereign wealth fund, established in 1990 to manage oil revenues. Worth around $2T, it funds public services via parliamentary-set payouts (though it’s not a direct citizen dividend or basic income).
Norway’s sovereign wealth fund was established in 1990 to manage the country’s North Sea oil revenues. It’s now the largest sovereign wealth fund in the world, worth roughly $2 trillion, and owns about 1.5% of every listed company on earth. The fund’s mandate is set by the Norwegian Parliament. Its ethical guidelines (no tobacco, no thermal coal, no companies that produce nuclear weapons) are publicly debated and revised. It publishes its holdings down to the share. It pays out a portion of its value annually to fund public services for Norwegians.
The fund has been criticized for being too large to invest neutrally, for the political fights over its ethical exclusions, and for its exposure to fossil fuel revenues that funded its creation. But it’s one of the most transparent investment vehicles on earth, and the Norwegian public has a claim on it that no court, no terms-of-service change, and no corporate restructuring can take away.
That is what a public version of “collective ownership” looks like. It’s audited in public, accountable to voters, and run for the citizens whose resources funded it.
What OpenAI says vs. what it funds
OpenAI’s April 2026 policy paper, “Industrial Policy for the Intelligence Age,” calls for higher capital gains taxes, taxes on “automated labor,” a Public Wealth Fund that would give every citizen a stake in AI-driven growth, portable benefits decoupled from employment, formal worker voice in AI deployment decisions, and a “Right to AI” framed as comparable to electricity and internet access.
On paper, it’s a progressive document.
In practice, OpenAI’s leaders are doing the opposite of what the document claims to support. OpenAI president Greg Brockman is funding a PAC called Leading the Future, dedicated to defeating state-level AI safety regulations, including some that OpenAI’s own white paper claims to endorse.
The same company calling for an expanded welfare state in PDFs is funding the political infrastructure attacking the welfare state in fights that are actually happening.
Joseph Stiglitz, the Nobel laureate economist, calls this part of a calculated strategy:
“The tech bros… are at the same time pushing for smaller government, which will undermine the ability of the government to do exactly what is needed in order to make a successful transition. The proposals on paper require a state strong enough to fund and administer them. The political spending is dismantling that state. The two moves work together."
Funding the AI social contract: Who pays?
Most of the tech industry seems happy to talk about basic income as long as the conversation doesn’t get too close to the part about taxing them to fund it. US Representative Bonnie Watson Coleman, who introduced basic-income legislation, told Semafor in April, “I don't think it's out of charity. It's out of business.”
Another US Representative Alexandria Ocasio-Cortez said: “I am skeptical about their willingness to pay or incur the taxes necessary to sustain such proposals, which would have to target AI.”
Canada could afford some form of basic income. But how? Higher marginal rates on top earners, a real digital services tax on AI firms, a windfall tax on the productivity gains AI generates, closing capital gains loopholes, and taxing automated labour the way human labour gets taxed. This is how a democracy funds a redistributive program - and each part of it is politically difficult.
A federal basic income would cost Canada roughly $107 billion gross annually, with a net federal cost of $3.6 billion to $5 billion after offsets, depending on whether the program replaces or stacks on top of existing benefits.
From Canada’s Parliamentary Budget Officer’s February 2025 estimate
An act to develop a national framework for a guaranteed livable basic income, Bill C-223, was defeated in late 2024. But public sentiment is moving. An August 2025 Abacus Data poll found that 32% of Canadians said a basic income guarantee was the most effective way to mitigate AI’s impact on jobs.
UBI’s designer makes a difference
The next time you read about Sam Altman’s latest UBI proposal or Elon Musk’s latest abundance pitch, the most useful question to ask is “Why is he the one designing it?”
Whichever way it goes, the people designing the program should be people we elected, and the money should come from the companies profiting from the AI transition.
AI in the news
Musk says he ‘was a fool’ to provide OpenAI’s early funding (NYT) Elon Musk is suing OpenAI and its partners, arguing the company abandoned its nonprofit mission by shifting to a for-profit model and commercializing AI, and is seeking $150B in damages. The case could determine whether AI companies are governed as public-interest organizations or profit-driven businesses.
OpenAI breaks off Microsoft exclusivity ahead of ChatGPT maker’s planned IPO (Globe and Mail) Microsoft and OpenAI have renegotiated their partnership, ending Microsoft’s exclusive rights to sell OpenAI’s models. This allows OpenAI to strike deals with companies like Amazon and Google.
‘Awkward and humiliating’: UK job hunters share frustration with AI interviews (Guardian) Nearly half of UK job seekers have experienced AI-led interviews, with many describing the process as awkward, impersonal, and even “humiliating.” About 30% abandoned applications because of it. While companies use AI interviews to handle volume, candidates say the one-way, automated format strips out human interaction and may disadvantage certain applicants, raising questions about fairness and effectiveness in hiring.
So basically Altman and Musk are proposing redistribution mechanisms that run through infrastructure they own? They're bypassing the democratic process so they can own the mechanism themselves, which honestly makes sense for rational actors when government is slow to react. I get it. But like you said....why does it have to be these guys? Problem is these are billionaires designing the safety net they'd otherwise owe the most under.
The only way I can see this actually working is if the government takes ownership stakes in key industries but Americans call that socialism. So the real fight is probably taxation. But whoever designs the mechanism first wins, and regular folks are not exactly in a position to out-maneuver that. Really interesting piece, Nicolle.
Thank you for covering this so thoroughly.
The convoluted altruism is the same defense I see across compliance work: propose a voluntary framework precisely when mandatory regulation starts looking inevitable. They are afraid of losing their special incentives and being taxed like any other business.