Bridgewater Associates is urging policymakers to consider a 35% tax on AI use, with revenue used to give the public a stake in AI companies and help workers. The proposal, reported by the New York Post, turns the debate over AI’s economic gains into a concrete question: who should receive them, and how?
Watch Desk analysis
What happened
Bridgewater executives Greg Jensen and Nir Bar Dea have published proposals for what the firm calls “citizen equity”. The Post reports that Bridgewater has also held talks with policymakers from both parties about the ideas.
One proposal would tax AI usage at 35%. Bridgewater estimates that could raise $600 billion by 2030, with the money used to acquire shares in AI companies for the public, cut taxes on human workers and support people whose jobs are displaced. The firm also proposes heightened oversight for AI companies controlling more than roughly 5% of US or global computing power, and more active government scrutiny of AI safety.
What changes
- A tax on AI use
Bridgewater proposes a 35% levy, arguing that machine labour should not be favoured over human work. - Public ownership of AI companies
The proposed revenue could buy shares for distribution to the public, rather than fund universal basic income. - More scrutiny of dominant firms
Jensen suggests companies above roughly 5% of AI computing power could face enhanced oversight. - Safety interviews for lab staff
Bridgewater proposes regulators conduct sworn interviews about emerging-model risks and the steps labs take to address them.
Why it matters
This is more than a call to help workers after disruption: public ownership would give people a direct financial stake in the companies benefiting from AI. Bridgewater argues that approach would put power in citizens’ hands, while critics quoted by the Post question whether buying AI shares is the right use of public money.
The idea also has a practical snag. Jensen acknowledges that collecting a tax on AI use would require a workable metering system, and the Post quotes an AI adviser pointing out that there is no universal standard for measuring AI usage. A large forecast is not the same thing as a tax that can actually be collected.
Our read
Bridgewater has put specific machinery on the table: a tax, public share ownership and tougher oversight. That makes the proposal worth debating, even if “35%” is the easy part and deciding what counts as taxable AI use is likely to be the rather less tidy one. Treat the $600 billion estimate as Bridgewater’s projection, not a settled revenue forecast.
What to watch
- Whether policymakers take up the proposal in formal legislation or hearings.
- How a tax would define and measure AI usage across different services.
- Whether public share ownership, worker tax cuts or direct support would win favour.
Discussion spark: If AI use were taxed, should the proceeds buy public shares in AI companies, reduce taxes on workers, or go directly to people displaced by automation?
Sources and evidence
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.