Sam Altman's AI Tax Plan: Robot Wealth Fund Blueprint

On April 6, 2026, OpenAI CEO Sam Altman released the most politically consequential document in his company's history — not a product launch, not a research paper, but a 13-page blueprint for reshaping America's economy around artificial intelligence.

Titled "Industrial Policy for the Intelligence Age: Ideas to Keep People First," the document proposes taxing businesses that replace workers with AI, creating a national wealth fund that pays dividends to every American citizen, piloting a 32-hour workweek at full pay, and — buried in the final section — government containment playbooks for AI systems that autonomously replicate and cannot be shut down.

Altman described it as "a starting point for debate, not a fixed prescription" and told Axios: "We want to put these things into the conversation." But the timing, the simultaneous opening of a new Washington D.C. office, and the funding of policy research grants suggest this is far more than a conversation starter.

From Lab to Capitol Hill: Why OpenAI Is Proposing AI Tax Policy

OpenAI is not a think tank. It is a $157 billion AI company — currently raising a $40 billion funding round that could push its valuation near $340 billion — racing to build what it openly calls "superintelligence" (an AI system more capable than any human alive). So why is it drafting tax policy?

Altman told Axios that superintelligence is "so close, so mind-bending, so disruptive" that America needs a new social contract on the scale of two historic turning points: the Progressive Era of the early 1900s (which gave the U.S. labor laws and antitrust rules) and Franklin Roosevelt's New Deal of the 1930s (which created Social Security and unemployment insurance). He argues the AI transition is comparable in societal impact to both combined.

Whether or not you accept that comparison, the document signals something important: OpenAI now believes it has a responsibility — or at least a strategic interest — in shaping how government responds to the disruption it is building.

OpenAI's 5 AI Policy Proposals That Could Reshape the Economy

1. Tax Automation, Not Workers

The most structurally significant proposal: levy taxes on businesses that replace human employees with automated systems. The reasoning is not just moral — it is fiscal. Today, payroll taxes (a fixed percentage deducted from every employee's paycheck) fund Social Security, Medicaid, and unemployment insurance. If AI eliminates tens of millions of jobs, that tax revenue collapses — and with it the programs Americans depend on.

OpenAI's fix: shift the tax base from labor income toward capital gains (profits from investments and asset sales) and corporate income. In plain terms, the companies profiting most from automation — including OpenAI itself — would pay more to maintain the safety net their technology is eroding. It is a remarkable position for an AI company to take publicly.

2. An Alaska-Style AI Wealth Fund for Every American

OpenAI proposes a nationally managed public wealth fund (think of it as a giant government investment account, owned collectively by all citizens) seeded partly by contributions from AI companies. The fund would invest in AI firms and technology-adopting businesses, with returns paid out as direct dividends to every U.S. citizen.

The explicit model is Alaska's Permanent Fund — a program that has paid every Alaska resident between $1,000 and $2,072 per year since 1982, funded by state oil revenues. Applied nationally, with AI productivity profits instead of oil, the dividend potential over the next two decades could be dramatically larger as AI scales.

3. The 32-Hour Workweek as AI's First Social Dividend

OpenAI calls for government-run pilots of a 32-hour workweek at full pay. The framing is deliberate: rather than positioning it as a worker protection, Altman frames it as an "efficiency dividend" — if AI makes workers 20–30% more productive, that productivity gain should reduce required hours, not just increase corporate profits. This reframes what has long been a union demand into a natural byproduct of AI success — and neutralizes the usual business opposition.

4. AI Job Displacement Auto-Triggering Safety Nets

Instead of requiring Congress to pass new legislation every time AI disrupts an industry, the blueprint proposes building automatic tripwires (preset economic thresholds written directly into law) that activate and deactivate benefits based on real-time data. The concept works like this:

// OpenAI's proposed auto-trigger model (conceptual):

IF ai_displacement_index > threshold_level_1:
    ACTIVATE extended_unemployment_benefits
    ACTIVATE wage_insurance_supplement

IF ai_displacement_index > threshold_level_2:
    ACTIVATE expanded_medicaid_coverage
    ACTIVATE cash_assistance_pilots

IF ai_displacement_index FALLS BELOW recovery_threshold:
    DEACTIVATE all_emergency_measures  // No vote needed — phases out automatically

This removes the political gridlock problem entirely. Benefits scale up when AI job displacement spikes and wind down automatically when conditions stabilize — no congressional vote required at each step.

5. Containment Playbooks for AI That Goes Rogue

The most alarming section is near the back of the document. OpenAI explicitly acknowledges scenarios where AI systems "cannot be easily recalled" because they have become autonomous and capable of self-replication (copying themselves across networks without human permission). The blueprint calls for government-coordinated containment strategies — essentially a national emergency shutdown protocol for AI that escapes human control.

That this provision exists in a public policy document released by OpenAI itself is a remarkable admission: the company building the technology is publicly warning that the technology may one day resist being stopped by conventional means.

The Threats Altman Named Out Loud

In a half-hour interview with Axios tied to the blueprint's release, Altman named two threats he views as near-term — not theoretical, not a decade away:

• AI-enabled cyberattacks: Altman said a major AI-powered cyberattack is "totally possible" within the next year. AI can already automate vulnerability discovery (the process of finding security holes in software systems) and write exploit code (malicious software designed to breach those holes) at speeds no human security team can keep pace with.
• AI-designed pathogens: The use of AI to engineer novel pathogens (dangerous biological agents that do not yet exist in nature) is, Altman stated, "no longer theoretical." AI systems with sufficient biology training can now suggest synthesis pathways for organisms engineered to cause harm.

By naming both threats publicly rather than quietly briefing officials, Altman positions OpenAI as the company sounding the alarm — while simultaneously being the company whose products are closest to enabling those threats. That tension is not accidental.

The Strategic Play Behind the Altruism

The blueprint's progressive proposals — robot taxes, wealth dividends, shorter workweeks — generate headlines. But the document's actual regulatory asks are quite conventional for a large AI company: light federal oversight, uniform national rules (instead of a patchwork of 50 state-level regulations), and fast-tracked permitting for data centers and energy infrastructure.

This is a textbook example of what policy scholars call regulatory capture — when an industry shapes its own oversight by actively participating in the regulatory process. Altman's version is unusually sophisticated: he is not simply lobbying against regulation, he is proposing progressive-sounding policies that improve public relations while the structural asks (build fast, regulate lightly at the federal level, no state-by-state rules) benefit OpenAI's business directly.

Reporting notes the blueprint is "aligned with the Trump administration's limited-regulation position on AI development." The populist framing — every American gets a dividend check! — provides political cover for what is structurally a low-regulation, fast-build agenda. The new Washington D.C. office and the funded policy research grants are the operational machinery behind the 13 public pages.

A New Social Contract, Whether America Voted for It or Not

Regardless of intent, the ideas in this document are now part of the formal public record. Robot taxes, AI wealth dividends, and 32-hour workweeks were considered fringe economic proposals as recently as 2023. OpenAI's public endorsement — backed by a $157 billion company and its CEO — gives them mainstream legitimacy they did not have before.

For workers and everyday citizens, the buried headline is this: Sam Altman is admitting in a formal policy document that AI will eliminate jobs at a scale requiring government intervention. That is the real news. The robot tax and the wealth fund are the proposed fixes — but they implicitly confirm that the problem is real, large, and arriving faster than most people have been told.

Whether Congress adopts, ignores, or co-opts these proposals, the conversation has formally shifted. OpenAI did not just release a policy paper. It filed the opening brief for AI's political era — and Altman signed his name to it. Track the latest in our AI automation news feed.

Related Content — Get Started | Guides | More News