OpenAI’s 5% US Government Stake: What Altman’s Wealth-Sharing Plan Actually Means

Sam Altman, CEO of OpenAI, is reportedly in talks with President Trump about giving the US government a 5% stake in OpenAI, according to a Financial Times report published on Thursday. The talks bring renewed attention to Altman's long-standing idea that ordinary Americans should receive a share of the wealth that AI companies generate. A per-household estimate puts that potential stake at around $300 per American family, though the final shape of any deal remains unclear.

What happened

The Financial Times reported on Thursday that OpenAI CEO Sam Altman is in active talks with President Trump’s administration about handing the US government a 5% ownership stake in OpenAI. MIT Technology Review first flagged the story through its weekly AI newsletter, The Algorithm.

This is not a brand-new idea from Altman. Back in 2021, he wrote publicly about a much broader version: requiring all companies above a set valuation threshold, not just AI companies, to contribute 2.5% of their market value each year into a national fund. That fund would then send annual payments directly to American citizens.

The current proposal is more limited in scope. In April 2025, OpenAI described a narrower version that appears to closely match what Altman is now negotiating with the White House.

Why it matters

OpenAI’s valuation has climbed sharply over recent years. A 5% government stake in a company at that scale translates, on a rough per-household basis, to something in the range of $300 per American family. That figure has circulated widely as a way to make the abstract idea feel concrete.

The proposal sits at the intersection of two pressures Altman has been navigating for years. First, OpenAI needs to keep regulators and policymakers on side as it pursues a conversion from nonprofit to for-profit structure. Second, Altman has consistently argued that AI will concentrate wealth rapidly, and that some mechanism is needed to distribute gains more broadly.

What form the stake would take matters a great deal. A passive government equity position is very different from a fund that sends cash to households. The 2021 proposal was the latter. The current talks appear to focus on the former, which means the “$300 per family” framing is more illustrative than literal, at least for now.

Our take

From where we sit, this story deserves more scrutiny than it usually gets.

Altman’s 2021 proposal was genuinely bold: a broad wealth tax on large companies, redistributed directly to citizens. What’s on the table now sounds much more like a political goodwill gesture. Giving the US government a 5% passive stake in OpenAI is not the same as sending Americans a check. It benefits the government’s balance sheet, not household bank accounts.

There’s also a structural question worth asking: why now? OpenAI is in the middle of a significant corporate restructuring. Keeping Washington favorable during that transition is a clear business interest. The timing of these talks is not coincidental.

None of this means the idea is worthless. A government stake could theoretically create accountability pressure that a purely private OpenAI would not face. But readers should be skeptical about how directly this translates to personal financial benefit without seeing the actual deal terms.

What to do about it

If you run a business that depends on OpenAI’s API or products, watch how this deal develops. A US government equity stake could eventually influence OpenAI’s pricing, data policies, or access terms, particularly if it comes with strings attached. Sign up for deal updates from reliable sources like the Financial Times or MIT Technology Review, and revisit your vendor diversification strategy if you are currently all-in on OpenAI infrastructure.

Source: MIT Technology Review