OpenAI Wants to Tax Itself. Robot Taxes, Free Money, and a 4-Day Week from the $840B Company.

A company valued at $840 billion just asked the government to tax it more. On April 6, 2026, OpenAI published a 13-page policy document titled Industrial Policy for the Intelligence Age that proposes robot taxes, a sovereign wealth fund seeded by AI companies, and government-backed 32-hour workweeks at full pay.

Sam Altman compared the scale of what is coming to the Progressive Era and the New Deal. Those were responses to economic crises that reshaped American society for a century. Altman thinks AI displacement will be that big. And he wants the government to start building the safety nets now.

Six Proposals That Would Reshape the American Economy

The document is not a vague manifesto. It contains six concrete policy proposals, each with mechanisms that could be legislated. Here is what OpenAI is actually asking for.

1. Robot Taxes

OpenAI proposes taxes on automated labor. The logic: if AI replaces human workers, the payroll taxes those workers paid into Social Security, Medicaid, and SNAP disappear. The federal government loses its funding base. Robot taxes would replace that lost revenue by taxing the machines that took the jobs.

The document calls for shifting the entire tax base away from wages and toward capital gains and corporate income. This is not a small adjustment. Payroll taxes generate roughly $1.6 trillion annually for the federal government. Replacing that revenue stream requires a fundamental restructuring of how America funds its safety net.

2. Public Wealth Fund

Every American would receive an equity stake in AI-driven economic growth through a nationally managed sovereign wealth fund. The fund would be seeded partly by AI companies themselves and would invest in "diversified, long-term assets that capture growth in both AI companies and the broader set of firms adopting and deploying AI."

Think of it as a national 401(k) that you do not have to contribute to. The government buys stakes in AI companies and distributes the returns to all citizens. Norway runs something similar with its $1.7 trillion Government Pension Fund. Alaska distributes oil revenue to residents through the Permanent Fund Dividend. OpenAI wants to do the same with AI profits.

3. 32-Hour Workweek at Full Pay

Government-backed pilot programs would test 32-hour workweeks while maintaining current pay levels. The idea: AI productivity gains should translate into more leisure time for workers, not just higher profits for shareholders. Pilots would involve employers and unions co-designing the programs.

The UK's four-day workweek trial in 2022 showed results. 61 companies participated. Revenue stayed flat or increased at 95% of them. Employee burnout dropped 71%. But that trial did not involve AI automation as the driver. OpenAI's proposal is different — it assumes AI handles the lost 8 hours of output.

4. Auto-Triggering Safety Nets

When AI-driven displacement hits preset thresholds, benefits increase automatically. Unemployment payments, wage insurance, and cash assistance would scale up without waiting for Congress to pass emergency legislation. When conditions stabilize, the extra benefits phase out.

This is the most pragmatic proposal in the document. Legislative responses to economic crises take months or years. Automatic stabilizers remove the political bottleneck. The current system already does this with unemployment insurance extensions during recessions. OpenAI wants to apply the same mechanism to AI displacement specifically.

5. AI Access as a Public Right

Affordable AI access would be treated as essential infrastructure — like electricity, water, and broadband. Schools, libraries, small businesses, and underserved communities would receive subsidized or free access to AI tools.

This reframes AI from a luxury product to a public utility. If AI becomes necessary for employment (as the internet already is), denying access creates a new class of economic exclusion. OpenAI is preemptively arguing for universal access before that exclusion solidifies.

6. Rogue AI Containment Playbooks

The document acknowledges scenarios where autonomous AI systems "cannot be easily recalled" because they are self-replicating. OpenAI's answer: coordinated government response frameworks developed in advance.

This is the most surprising inclusion. A company that sells AI is explicitly warning that future AI systems might be uncontrollable. The containment playbook proposal suggests OpenAI believes this scenario is plausible enough to plan for publicly.

The Hypocrisy Question: Can a $840B Company Ask for Higher Taxes?

OpenAI closed a $110 billion funding round in February 2026 at an $840 billion post-money valuation. The company runs $25 billion in annualized revenue. It is planning an IPO for Q4 2026 or Q1 2027, potentially targeting a $1 trillion market cap.

And it just asked the government to raise capital gains taxes on AI company profits.

The cynical reading: OpenAI is trying to look responsible before its IPO. A company calling for its own taxation makes great PR. It costs nothing until legislation actually passes — and OpenAI's army of lobbyists can shape the details when it does.

The charitable reading: OpenAI genuinely believes AI displacement will be catastrophic and is trying to build political support for safety nets before the crisis hits. Sam Altman has been writing about universal basic income since 2016. This document is the policy version of that belief.

The realistic reading is probably somewhere in the middle. OpenAI benefits from regulation that it helps write. A Public Wealth Fund seeded by AI companies sounds egalitarian until you realize the largest AI company gets the most influence over fund governance. Robot taxes sound progressive until the tax code's implementation details favor companies with the most lobbying budget.

How This Compares to OpenAI's January 2025 Blueprint

In January 2025, OpenAI published an "Economic Blueprint" focused on infrastructure investment and light-touch regulation. That document proposed AI economic zones, a "National AI Infrastructure Highway" for power and communications grids, and faster permitting for AI data centers. The total investment opportunity cited was $175 billion in global funds waiting for AI projects.

The 2025 blueprint read like a corporate wish list. Build us infrastructure. Cut our red tape. Let us grow.

The 2026 document reads like a social contract. Tax us. Distribute the wealth. Protect the workers we displace.

The shift in 15 months is dramatic. Something changed inside OpenAI's policy team. The most likely explanation: internal data on AI capabilities convinced leadership that displacement will happen faster than the 2025 blueprint assumed. When you believe your product will eliminate millions of jobs, "build more data centers" stops being a sufficient policy position.

Anthropic Just Passed OpenAI in Revenue

While OpenAI was publishing policy papers, Anthropic quietly crossed $30 billion in annualized revenue — surpassing OpenAI's $25 billion for the first time. Anthropic's growth rate is roughly 10x per year, versus OpenAI's 3.4x.

Anthropic is also evaluating an October 2026 IPO at a potential $380 billion valuation. The AI industry now has two companies approaching trillion-dollar valuations simultaneously.

This context matters for the policy document. OpenAI is not the only company whose profits would fund a Public Wealth Fund. Anthropic, Google, Meta, and every other AI company would contribute. OpenAI calling for AI taxes is partly calling for taxes on its competitors too.

For a look at how AI companies are building the tools that might displace workers, check out Glide — a no-code AI app builder that lets non-developers create business applications from spreadsheets. And for the open-source infrastructure powering AI agents, explore Vercel's agent-browser, which gives AI systems direct browser control.

What Happens Next

Nothing immediately. This is a policy proposal, not legislation. No member of Congress has introduced a bill based on the document. The Trump Administration's National AI Legislative Framework from March 20, 2026, focuses on industry-friendly regulation, not robot taxes.

But the Overton window just shifted. The largest AI company in the world publicly stated that AI will displace enough workers to collapse the tax base. That admission from inside the industry carries weight that external critics cannot match.

Three things to watch:

Congressional hearings. Expect Sam Altman to testify within 90 days. The document gives lawmakers a specific agenda to question him about. "You said AI will collapse Social Security funding. When? How many jobs?"

Competitor responses. Anthropic, Google, and Meta have not responded publicly. Their silence is temporary. Each will need to position itself on robot taxes, wealth funds, and workweek regulation. Staying silent looks like opposition.

The IPO calculus. OpenAI's $14 billion projected loss in 2026 makes the IPO narrative tricky. A company losing money while asking for higher taxes needs a compelling story. The policy document might be the beginning of that story — "we are not just profitable, we are responsible."

For more on how AI governance tools are evolving alongside these policy debates, explore the MCP Gateway Registry for enterprise-grade AI agent management, and browse the latest AI tools directory on Skila.

The Bottom Line

OpenAI just told the world that AI will be so disruptive it needs a New Deal-scale response. Robot taxes. A national wealth fund. Four-day workweeks. Automatic safety nets. Containment plans for rogue AI.

Whether you believe OpenAI is sincere or strategic, the proposals are now part of the policy conversation. A $840 billion company does not publish a 13-page document calling for its own taxation without expecting consequences. The question is whether those consequences include actual legislation — or just better IPO optics.

Frequently Asked Questions

What is OpenAI's Industrial Policy for the Intelligence Age?

It is a 13-page policy document published April 6, 2026, proposing six economic reforms to address AI-driven worker displacement. The proposals include robot taxes, a Public Wealth Fund giving all Americans equity stakes in AI growth, 32-hour workweeks at full pay, auto-triggering safety nets, universal AI access, and rogue AI containment playbooks.

How would a robot tax work under OpenAI's proposal?

OpenAI proposes taxing automated labor to replace the payroll tax revenue lost when AI displaces human workers. The tax base would shift from wages toward capital gains and corporate income. Payroll taxes currently generate roughly $1.6 trillion annually for Social Security, Medicaid, and SNAP. Robot taxes would fill that funding gap.

How does this compare to OpenAI's January 2025 Economic Blueprint?

The January 2025 blueprint focused on building AI infrastructure — data centers, power grids, economic zones, and faster permitting. The April 2026 document shifts dramatically toward social protection: taxing AI profits, distributing wealth, and protecting displaced workers. The change suggests OpenAI internally revised its estimates of how fast AI displacement will occur.

Is OpenAI more valuable than Anthropic in 2026?

OpenAI's post-money valuation is $840 billion after its February 2026 funding round, with $25 billion annualized revenue. Anthropic surpassed OpenAI in annualized revenue in April 2026, hitting $30 billion with roughly 10x year-over-year growth versus OpenAI's 3.4x. Anthropic's projected IPO valuation is $380 billion.

Will OpenAI's robot tax proposal become law?

Not immediately. No legislation has been introduced based on the document. The Trump Administration's March 2026 AI framework favors industry-friendly regulation. But the proposal shifts the policy conversation — expect Congressional hearings and competitor responses within 90 days.