Bernie Sanders’ AI sovereign wealth fund plan is good. But we think this is better
The authors back Sanders’ goal of curbing AI power concentration, but argue public stock ownership would create perverse incentives. They prefer taxes and publicly run AI models instead.
Intelligence analysis by GPT-5.4 Mini
Nathan E Sanders and Bruce Schneier agree AI is concentrating wealth and power, but they reject the idea that the government should own large stakes in AI firms. They argue that taxes can share the gains, while a public AI option would better keep the technology under democratic control.
The writers think AI is making a few people very rich, like one kid getting all the candy. But instead of the government owning part of those candy factories, they want taxes to share the money and a public AI helper that everyone can trust.
Analysis
The core disagreement
The authors say Bernie Sanders is asking the right question: whether the future will be shaped by a small set of AI billionaires with little democratic input. They agree that AI is concentrating wealth and power and that governments should act to protect the public interest.
Why they reject the sovereign wealth fund idea
Sanders has proposed taking a 50% stock stake in major AI companies such as Anthropic, OpenAI, and xAI to create a U.S. sovereign wealth fund. The authors say that could give the public a share of AI’s financial upside and some governance influence. But they argue that public ownership would also tie the government’s interests to the companies’ valuation.
In their view, that creates the wrong incentive structure. If public officials benefit when AI firms grow, they may be pushed to weaken regulation, tolerate worker and user harms, and encourage AI adoption even when the use case is not responsible or appropriate. The piece argues that this would make corporate influence on government more likely, not less.
What they think works better
The authors separate the two goals Sanders is trying to achieve. To share economic gains, they prefer taxation, including ideas like an excise tax on data centers’ energy use or an AI token tax. To shape AI in the public interest, they propose a public option: government-developed and government-run AI models under democratic control.
They point to Switzerland’s public-sector model, Apertus, as an example of this approach. Their broader argument is that the state should not own private AI firms in order to influence them; it should instead tax them and build public alternatives.
Key points
- The authors agree with Sanders that AI is concentrating power and wealth in too few hands.
- They support the goals of public influence and public benefit, but not public stock ownership in AI firms.
- Their main objection is that owning shares would tie government incentives to corporate valuation.
- They favor taxes, such as an energy tax on data centers or an AI token tax, to share AI gains with society.
- They propose a public AI option as a better way to create democratic control and competition.
If taxes on AI infrastructure or usage are adopted, some of the money flowing to AI companies could be redirected to the public. A government-built AI option could also give people and businesses a non-corporate choice under democratic control.
If the government owns big stakes in AI companies, it may be tempted to protect their profits instead of regulating them hard. That could weaken oversight, encourage reckless AI adoption, and increase the influence of corporate interests over public policy.
