Senator Bernie Sanders has unveiled legislation that would mandate 50% public ownership of major AI companies through a sovereign wealth fund, representing the most ambitious attempt yet to reshape the AI industry's ownership structure and ensure its benefits flow to the public.
The proposal, which would apply to the largest AI firms including OpenAI, Anthropic, Google DeepMind, and others, would create an AI sovereign wealth fund that holds direct equity stakes in these companies. Sanders argues that since AI development has relied heavily on publicly-funded research and infrastructure, the public deserves to share in the returns.
From a policy perspective, it's an interesting idea. Much of the foundational AI research was publicly funded. The transformer architecture came out of Google, but built on decades of academic research supported by government grants. The internet infrastructure these companies rely on was built with public investment. Even the electricity grids powering their data centers often benefit from public subsidies.
The tech industry's response has been predictably hostile. Industry groups argue that public ownership would stifle innovation, claiming that private capital and competition drive AI advancement. But that narrative conveniently ignores how much AI companies already depend on public resources - from research to infrastructure to the legal frameworks that allow them to train on copyrighted data.
What's less clear is whether the proposal is practically workable. How would you value these stakes? Would the government have voting rights or just economic interests? What happens when companies make decisions that serve public rather than private interests - like prioritizing safety over speed to market?
Norway's sovereign wealth fund, built on oil revenues, provides one model. Alaska's Permanent Fund provides another. Both distribute resource wealth to citizens rather than letting it concentrate in private hands. Sanders is essentially arguing that AI represents a similar natural resource - one created through collective investment that shouldn't be privately captured.
The proposal has approximately zero chance of passing in the current political environment. But it's shifted the conversation. Six months ago, suggesting public ownership of AI companies would have been dismissed as fringe. Now it's a serious policy proposal being debated in congressional hearings.
