AI Belongs to No One — or Perhaps to Everyone: The Battle That Will Shape the Future

Bernie Sanders, the Vermont senator who for decades has embodied the most progressive wing of American politics, has introduced a bill that is shaking Silicon Valley to its core. The objective is clear and brutal in its simplicity: to force giants like OpenAI, Anthropic, and Elon Musk's xAI to hand over fifty percent of their shares to a public fund controlled by the federal government of the United States. Not a tax, not a regulation. Half the company. Full stop.

The immediate reaction from those who follow technology markets was predictable: cries of alarm, accusations of socialism, comparisons to Venezuela. But stopping at the surface of this proposal would be a mistake. Because the reasoning Sanders advances has an internal logic that is difficult to dismantle with the usual free-market arguments.

Artificial intelligence, as we know it today, was trained on a quantity of data that defies imagination. Books, articles, source code, social media posts, online conversations, literary, musical, and artistic works. Decades of humanity's collective intellectual output, generated in the vast majority of cases by people who never received a single cent for that contribution. Large Language Models like GPT, Claude, and Grok literally devoured the collective mind of human civilization to build their systems. And the profits from this operation — astronomical profits, according to Goldman Sachs projections estimating the global AI market at over 1.8 trillion dollars by 2030 — would end up in the pockets of a small group of billionaires. Sam Altman, Elon Musk, Dario Amodei. Brilliant people, no doubt, but very few of them.

This is the short circuit that Sanders wants to break with his proposal. It is not an abstract ideological matter. It is a very concrete question: if the engine of artificial intelligence was built using the unpaid intellectual labor of billions of human beings, who truly has the right to the profits that derive from it?

Silicon Valley's answer is equally concrete: it was private capital, entrepreneurial risk, and the strategic vision of the founders that transformed that raw material into functional technology. Without billions in private investment, without years of expensive research, that data would have remained just data. It is capital, not the crowd, that created value.

Both positions contain a grain of truth. And it is precisely for this reason that Sanders' proposal, however extreme it may seem right now, represents something more than a mere electoral stunt. It is the first visible signal of a tension that will inevitably grow in the coming years, as artificial intelligence penetrates every sector of the economy, redefines the labor market, and concentrates unprecedented historical power in the hands of those who control the most advanced models.

The International Monetary Fund, in its January 2024 report titled AI and the Future of Work, estimated that artificial intelligence could impact sixty percent of jobs in advanced economies. Not necessarily eliminate them, but transform them radically. In this scenario, the question of who controls AI stops being philosophical and becomes urgent, practical, and political in the most literal sense of the word.

The real war opening up is not between Sanders and Altman. It is between three incompatible visions of the future: AI as private corporate property, AI as a public good controlled by states, and AI as open, distributed infrastructure that belongs to no one in particular. The Vermont senator's proposal is only the first official shot in a conflict destined to last decades and to redraw the global balance of power in ways we can today only begin to imagine.