April 2026 — Just today, OpenAI released a 13-page document they’re calling Industrial Policy for the Intelligence Age: Ideas to Keep People First. It’s a lot of things at once: a policy wishlist, a manifesto, and maybe a bit of a distraction. They’re arguing for things like public wealth funds, portable benefits for workers, and a 32-hour workweek. They even talk about a "Right to AI" that would be as fundamental as being able to read or having the lights turned on. Throughout the paper, they sound genuinely worried about the idea of wealth and power being held by just a few massive firms.
But you have to look at the calendar to see the real context. Only six days ago, OpenAI wrapped up a $122 billion private funding round. That deal gave the company a valuation of $852 billion—the kind of number we’ve never really seen before in Silicon Valley’s history of private financing.
If these two things happening in the same week feels like a strange coincidence, that’s because it isn't one. The timing is basically the whole story.
The Manifesto
From a high level, the document is put together well. OpenAI suggests three main ways to handle the transition to superintelligence: sharing the wealth, managing the various risks, and making sure access is actually democratic. They keep pointing back to the Progressive Era and the New Deal for inspiration. They’re also pushing for adaptive safety nets—systems that would kick in automatically if the labor market starts to break—and they think advanced AI companies should become Public Benefit Corporations that give away a chunk of their profits.
There’s a certain cleverness in the writing here. It acknowledges the risks that, ironically, make the company look even more important. They say there’s a risk that economic gains could end up "concentrated within a small number of firms like OpenAI." It’s an honest admission, but it’s also the most revealing part of the document. It was written by the very people who are building the machine they’re warning us about.
Beyond the big ideas, they’re setting up $100,000 fellowships and a workshop in Washington for May. They even have an email address set up for public feedback. But they’re very clear about one thing: they think this is just the beginning of a much longer conversation.
The Spreadsheet
In the real world of finance, the recent funding round was anchored by big names: Amazon, Nvidia, and SoftBank. Amazon’s $50 billion check actually has some strings attached—it’s contingent on the company going public or hitting some big AGI milestone. Sam Altman is already aiming for a Nasdaq listing by the end of the year, with the initial S-1 filings likely showing up this summer. They’ve been spending months building the formal structure a public company needs, from hiring a top accounting officer to beefing up their investor relations team.
But if you look at the accounts coming from their own CFO, the situation looks a lot more complicated. Sarah Friar has reportedly been telling people inside the company that OpenAI is carrying $600 billion in compute commitments over the five years ahead. Most of that is for the massive data centers they’re building with Amazon, Oracle, and Nvidia. She thinks they’ll burn through $200 billion in cash before they see any profit at all. They’re looking at a $14 billion loss just this year, and profitability isn’t even on the horizon until 2030. Their margins are tightening, and their grip on the market isn't what it used to be. Their share of the enterprise API market was basically cut in half while Anthropic’s share jumped from 12% to 32%.
The oddest part of this is how Friar is being treated. Reports suggest Altman has been keeping her out of meetings with the biggest investors because her realistic views on the cash burn might kill the momentum. They even changed her reporting line. In the tech world, that’s usually a signal that things are getting messy behind the scenes right before an IPO.
The Real Gap
The OpenAI document talks a lot about stopping "insider capture" and making sure no small group is able to hoard power over the technology. These are valid points, but they also happen to be a pretty accurate description of how OpenAI is being run right now. You have a CFO being pushed to the side for telling the truth about the money, a valuation that is entirely built on keeping their best technology rare, and an IPO timeline that is clearly being driven by the fear that a competitor like Anthropic will list first and take all the available cash from the public markets.
I don’t think this is necessarily a case of simple corruption. It’s more of a structural reality. OpenAI is doing exactly what you’d expect a rational player to do in a system that rewards monopolies and makes true altruism very expensive. The "Public Wealth Fund" is a great idea for a white paper, but that $852 billion valuation—which would shrink everyone else’s share of the pie before a fund is even started—is the reality of the deal they just signed.
At the end of the day, the document doesn't really explain how they can promise to democratize these tools when their entire business model depends on them staying expensive and exclusive. If we actually had a superintelligence that was "people-first," the company’s valuation would probably collapse overnight. They might mean well with the policy suggestions, but the math is what’s actually in charge.
The conversation they’re inviting the rest of us to join is one where the terms—and that $852 billion price tag—have already been decided in the dark.
not investment advice