OpenAI just published a policy document that reads like it was co-authored by Bernie Sanders and Milton Friedman. And that contradiction is the whole story.

The document, titled "Industrial Policy for the Intelligence Age," calls for public wealth funds so every American gets a stake in AI-driven prosperity. It proposes fast-response social safety nets for workers displaced by automation. It floats a robot tax, the idea Bill Gates pitched in 2017, where machines pay into the system the same amount as the humans they replaced. It even suggests subsidizing a four-day work week with no loss in pay.

If you are thinking "that sounds surprisingly progressive for the company whose president donated millions to Trump," you are not alone.

OpenAI president Greg Brockman and other tech billionaires have funneled hundreds of millions into super PACs supporting light-touch AI regulation. The Trump administration dropped the corporate tax rate to 21% from 35%. OpenAI's own document suggests raising taxes on corporate income and capital gains to fund AI transition programs. These positions are not compatible.

But that is precisely the point. This document is not policy. It is positioning. OpenAI is worth $852 billion and heading toward an IPO. It needs to look like it cares about the people its technology might displace, while simultaneously lobbying for the deregulated environment that lets it move fast. The wealth fund proposal is an insurance policy against the political backlash that is already building.

To OpenAI's credit, some proposals are genuinely smart. Shifting the tax base from labor to capital makes mathematical sense when AI replaces workers but grows corporate profits. The warning that AI could hollow out Social Security, Medicaid and SNAP funding as payroll tax revenue shrinks is something policymakers need to hear. The four-day work week idea aligns with JPMorgan CEO Jamie Dimon's recent prediction that AI will cut work to 3.5 days.

The problem is credibility. OpenAI is asking the government to build safety nets while it aggressively builds the technology that makes those safety nets necessary. It is proposing wealth redistribution while pursuing a valuation that would make it one of the most valuable companies in history. It is calling for grid modernization while planning data centers that will consume more electricity than small countries.

None of this makes the proposals wrong. A public wealth fund modeled on Alaska's oil dividend could genuinely help. Robot taxes could preserve public services. Shorter work weeks might be the best response to productivity gains that mostly benefit shareholders.

But when the company proposing these ideas is simultaneously spending millions to elect politicians who oppose them, the document reads less like industrial policy and more like a PR strategy with a really good bibliography.