OpenAI's $300-per-family idea reopens the AI wealth-share debate

The News

Sam Altman's long-running pledge that ordinary people should share in the wealth artificial intelligence creates is back in circulation, this time with a concrete price tag attached. According to a Financial Times report last Thursday, the OpenAI chief executive has been discussing handing the US government a 5% stake in the company.

Run the arithmetic and the headline shrinks fast. Spread that 5% holding equally across the roughly 133 million households in the United States, and each family would be sitting on about $320 of OpenAI equity. The figure is a useful reality check on a promise that has been debated, in various forms, for five years without producing much in the way of policy.

The underlying company is anything but small. OpenAI's March funding round valued the business at $852 billion, which puts the notional worth of a 5% slice at around $42.6 billion. The firm is reportedly holding back its public listing until it can command a $1 trillion valuation.

Why It Matters

Wealth-sharing schemes have followed Altman around for years. A 2021 version of the idea proposed that every company above a certain valuation pay 2.5% of its market value each year into a fund that would be disbursed to Americans. The current conversation, tied to talks with President Trump, is narrower and more transactional: a direct equity stake rather than a broad levy on the sector.

The politics sit on a spectrum. Senator Bernie Sanders has floated giving the public a 50% stake in the largest AI firms, a demand that makes Altman's 5% look modest by comparison. What all these proposals share is a recognition of two uncomfortable facts: AI models are built on human-generated work that is rarely paid for, and the same systems threaten to disrupt the labour markets that most households depend on.

History suggests caution. The last technology boom to promise mass shared ownership, the dot-com era's employee stock craze, ended with far more paper wealth than realised gains. Until OpenAI actually lists and its shares trade, a $320 stake is a narrative device, not a dividend.

Indian Angle

For India, the interesting part is not the American household cheque but the principle behind it: that a government can and perhaps should hold equity in the AI champions built on public data and public concern. India is already spending real money on the other side of that question. The IndiaAI Mission, backed by more than 10,000 crore rupees, is subsidising shared compute and backing domestic model builders such as Sarvam and Krutrim rather than taking stakes in foreign labs.

That is a deliberate choice. Where Washington debates clawing back a slice of a private giant, New Delhi is trying to manufacture sovereign alternatives from scratch, so the upside stays domestic by design. At today's exchange rate, $320 is close to 27,000 rupees, more than many Indian households would see from a subsidised compute programme, which sharpens the appeal of the ownership model for policymakers watching from Delhi.

There is a regulatory thread too. If governments start taking equity in AI firms, SEBI and MeitY will have to think about conflicts of interest, disclosure and how a state that both regulates and owns an AI company keeps the two roles apart. India's disinvestment history, from Maruti to LIC, offers plenty of lessons on how messy government ownership of strategic assets can become.

FAQ

How much would each US family actually get?

Based on a 5% OpenAI stake spread across roughly 133 million American households, about $320 of equity per family. That is a notional figure tied to OpenAI's $852 billion March valuation, not cash in hand, and only crystallises if and when the company lists.

Is this a done deal?

No. The Financial Times reported that Altman has been discussing a 5% US government stake, but nothing has been formalised. Similar wealth-sharing ideas have circulated for five years without becoming policy.

How does India approach the same problem?

Rather than taking stakes in foreign AI firms, India is funding domestic capability through the IndiaAI Mission, worth over 10,000 crore rupees, backing shared compute and home-grown model builders like Sarvam and Krutrim.

Where can I read the original report?

The analysis was published by MIT Technology Review, drawing on a Financial Times report about Altman's discussions with the US government.

This story was reported by MIT Technology Review. Read the full original coverage at MIT Technology Review.