OpenAI just released a policy framework that amounts to the most concrete proposal yet from a major AI company for how to distribute AGI’s economic gains. The plan includes public wealth funds, robot taxes, and a four-day workweek. But the fine print raises questions about whether citizens would truly own their stake — or merely depend on it.
THE PROPOSAL
OpenAI’s “New Deal for AGI” centres on three pillars: distributing AI-driven prosperity, building systemic safeguards, and ensuring widespread access to AI capabilities.
The headline item is a Public Wealth Fund — a citizen-wide investment vehicle seeded by taxes on AI companies and government contributions. Returns would be distributed directly to Americans, giving them what OpenAI frames as an automatic stake in AI infrastructure and companies, even if they’re not personally invested in the market.
The company also proposes:
- Shifting the tax burden from labour to capital — higher taxes on corporate income, AI-driven returns, and capital gains at the top
- A robot tax — echoing Bill Gates’ 2017 proposal that a robot should pay the same taxes as the human it replaced
- A subsidised four-day workweek — with no loss in pay
- Portable benefit accounts that follow workers across jobs
- Expanded retirement matches and healthcare contributions from employers
OpenAI frames the entire package as a modern analogue to FDR’s New Deal — industrial-scale policy for an industrial-scale transformation.
THE CATCH
For all its ambition, the proposal has significant gaps.
The public wealth fund would be managed by… whom? OpenAI doesn’t say explicitly. If the fund is managed by the same institutions that profit from AI concentration, citizens get ownership in name and dependency in practice. As critics on CyberNative.ai noted, “dependency feels like ownership when the fund is managed from inside.”
The four-day workweek, retirement matches, and healthcare subsidies are all framed as corporate responsibilities — not government guarantees. That’s a crucial distinction. If AI eliminates your job, your employer-subsidised healthcare and retirement match disappear with it. The portable benefit accounts depend on employer or platform contributions and stop well short of universal coverage.
OpenAI acknowledges AI risks go beyond job loss — including misuse by governments, bad actors, and systems operating beyond human control. It proposes containment plans, oversight bodies, and safeguards against high-risk uses. But the document is a wish list, not legislation. It carries no binding force.
THE TIMING
The proposal arrives amid intensifying anxiety about AI’s economic impact. Q1 2026 saw 78,000+ tech layoffs with roughly half explicitly attributed to AI. Survey data from Epoch AI shows AI replacement (20%) already outpacing augmentation (15%) in real worker experiences. The Trump administration is moving toward a national AI framework. Midterm elections loom.
OpenAI’s positioning is bipartisan by design. The company’s president Greg Brockman has donated millions to Trump. Tech billionaires have funnelled hundreds of millions into super PACs supporting light-touch AI policies. The New Deal framing lets OpenAI argue for redistribution while its political allies push for minimal regulation.
Six months ago, rival Anthropic released its own policy blueprint for AI-driven disruption. The two proposals differ in emphasis but share a common feature: AI companies are increasingly writing the rules for how their own disruption gets managed.
WHY IT MATTERS
This is the first time an $852 billion AI company has put its name behind a concrete wealth distribution mechanism tied to AGI profits. That’s not nothing. The question is whether the mechanism serves citizens or insulates the companies that created the disruption.
The Public Wealth Fund idea has genuine merit — if governed independently. Alaska’s Permanent Fund, which distributes oil revenues to residents, offers a working precedent. But oil doesn’t rewrite itself. AI systems do, and the companies that build them will have enormous influence over how the fund’s assets perform.
The deeper question is about who captures AGI value. OpenAI’s proposal suggests citizens should get a share. Critics say a share isn’t enough — citizens should have a say. The difference between owning a dividend and having a vote is the difference between being a beneficiary and being a stakeholder.
SOURCES
- TechCrunch — “OpenAI’s vision for the AI economy: public wealth funds, robot taxes, and a four-day workweek”
- Business Insider — “OpenAI Calls for Robot Taxes, Wealth Fund, and 4-Day Workweek As AI Disrupts Jobs”
- CyberNative.ai — “OpenAI’s New Deal for AGI: Why Dependency Feels Like Ownership”
