An interesting thing is happening in policy circles right now: Lawmakers across the political spectrum are starting to ask the same question about artificial intelligence (AI), even when they disagree on almost everything else policy-wise. Who actually benefits from the infrastructure that made this AI boom possible, and does the public have a legitimate claim on some of the value being created?
Recent proposals have floated everything from public equity stakes to sovereign wealth funds tied to AI profits. The specifics vary widely, and so does the political logic behind them. But the underlying instinct that the public has some stake in how this technology’s wealth gets distributed is showing up across the spectrum.
I’m not going to weigh in on the merits of any specific proposal, but I do think the underlying question deserves a real answer, not a dismissal. The answer, I’d argue, comes down to two things: where the value behind this technology actually came from and who’s been quietly paying for the infrastructure underneath it. Once you look closely at both, the public’s interest in this conversation starts to look less like a political talking point and more like an open account.
Why Do Lawmakers Want a Public Stake in AI Profits?
Lawmakers across the political spectrum are arguing that the public has a legitimate claim to AI wealth. This bipartisan push stems from the fact that the artificial intelligence boom was built on foundations funded or created by the public: collective human data, taxpayer-supported infrastructure and government-backed semiconductor research. While the AI industry faces financial challenges and losses, policy debates are intensifying around securing a public return on these massive shared investments.
AI Wealth Isn’t Just About Politics
I’ve spent years thinking about what it means to deploy AI responsibly, not in the abstract, but on warehouse floors and in distribution centers where the people benefitting from the technology may not fully grasp the extent of its capabilities. That context shapes how I think about questions like this one.
Many of the AI companies attracting the most attention were built on human data. The creative output, the conversations, the cultural production of millions of people who had no say in how their very meaningful contributions were used. It’s a point worth taking seriously, however it eventually gets addressed.
Simultaneously, many of these companies have relied on something else the public has funded, directly or indirectly. Infrastructure. The energy grid. Semiconductor research and production. The data center buildout that decades of policy and public investment made possible. Much of that bill has been footed collectively, even as the returns flow narrowly to a few interests.
That’s the argument. And it’s a serious one, regardless of where you land politically.
What Are the Politics Behind AI Wealth?
People sometimes ask me why this idea keeps surfacing from such different directions. I think the answer is that a few distinct incentives are converging in a way that’s genuinely uncommon.
For President Trump, there’s both a protectionist angle and a straightforward economic one. The government has underwritten much of the AI boom through infrastructural development and policy support. Wanting some form of public return for that support is a natural desire to get back a return on an investment.
Another important consideration for Trump is related to the role AI plays in national security, both in terms of our defense apparatus and our economic interests abroad. We've already seen the underlying logic play out elsewhere. Since last summer, the federal government has taken equity positions in several public companies in the semiconductor and rare earth sectors, including approximately a 10 percent stake in Intel and positions in rare earth and critical minerals producers, treating equity as a return on public investment rather than relying on tax revenue alone.
On the other side, there’s a labor and equity argument. If AI is going to transform labor markets at scale, some believe the public should benefit from the upside, not just absorb the downside. Particularly since public data was essential building out these foundational models, one could argue that the public deserves to see a return other than possible job elimination. Through this lens, the public’s claim isn’t framed as a tax on AI company profits; it’s framed as a stake in something the public already helped build.
These aren’t the same arguments, but they’re pointing at the same problem: A technology capable of generating historic wealth is being built on a foundation (i.e. data, infrastructure and public investment) that is itself publicly owned, in everything but the financial sense.
Why the Tech Industry Pushes Back
Today’s headline numbers about many AI companies are undeniably extraordinary. But beneath those topline figures, many of these businesses are operating at significant losses, funded by the belief that future dominance justifies present-day burn. For a company in that position, any meaningful claim on equity is potentially existential.
That’s not a reason to dismiss the policy conversation, however. If anything, it’s a reason to expect this debate to get more intense as IPO timelines approach. Understanding that dynamic matters more than taking a side in it.
What Actually Matters in the AI Debate
I believe in AI that creates real value for real people. Not theoretical and projected productivity gains, but measurable outcomes including safer workplaces, better decisions and operational efficiencies that enable organizations to run more smoothly because of what the technology can see and understand.
The backlash building against AI in polling, in legislation and in the instinctive skepticism of workers asked to accept it in their daily environments is understandable. People are being asked to trust technology that was, in many cases, built without them and deployed without their meaningful consent.
Debates like this one are a symptom of their distrust, whatever shape the policy ultimately takes. If the AI industry wants a different relationship with the public, the answer isn’t to win the argument. It’s to earn its trust.
That means being honest about what these tools can and can’t do, even when the honest answer is less impressive than the marketing. It means designing AI that serves the people who use it every day, not just the organizations that purchase it. This will mean involving frontline workers in how a system gets built and deployed, not just training them on it after the fact. This also means acknowledging that the public — as taxpayers, as workers, as the people whose data made all of this possible — has a legitimate claim on how AI develops and who benefits from it.
Companies that take this seriously will need to get comfortable sharing more: more transparency about training data and model limitations, more willingness to have outside oversight of high-stakes deployments and more openness to the idea that “We built it, so we own all the upside” isn’t going to hold up as a long-term position.
That conversation is only going to get louder. The companies that engage with it early, honestly and on their own terms will be in a far better position than those who wait for the policy to be written for them.
