A common argument for government regulation and taxation of tech companies begins with the observation that government research has contributed to the success of those companies. Mariana Mazzucato, professor of the Economics of Innovation and Public Value at University College London, writes:
Much of modern technology came from a collective investment, with public institutions like the US Defense Advanced Research Projects Agency (DARPA) or the European Council for Nuclear Research (CERN), leading the way in the most high-risk capital-intensive phase. What would Google be without the DARPA-funded internet? What would Uber be without the US Navy-funded GPS? What would Apple be without the CIA-funded touch-screen technology and DARPA-funded voice assistant, Siri?
Thinkers like Mazzucato seem to view the state as a sort of partner in these businesses; after all, it’s only fair that a partner should get a cut of the profits and a say in their operations.
“We need to develop a new governance structure, which starts with creating a new vocabulary,” she says. “For example, calling platform companies ‘tech giants’ implies they have invested in the technologies from which they are profiting, when it was really taxpayers who funded the key underlying technologies — from the Internet to GPS.”
This argument is particularly relevant to the case of AI, especially as both Donald Trump and Bernie Sanders have expressed support for partial government ownership of major AI companies. Mazzucato herself suggests that government officials ought “to actively steer [AI’s] development.”
At first glance, the claim that governments are entitled to help control AI companies because they helped create AI sounds plausible. If you were part of building something, don’t you deserve a share of it?
The case for government control of tech companies is far more complicated than it appears, however.
First, note that people are compelled to fund government research. To see why this makes a moral difference, consider a case inspired by the philosopher Michael Huemer.
Suppose you’re an entrepreneur in the beverage industry trying to create a new soda. One day, a rogue food scientist bursts into your office and demands money for his flavor research. He doesn’t ask politely or offer you a contract; he points a gun at you and takes your money. Years later, he develops a delicious new synthetic flavor. You discover it, add it to your soda, and your company takes off, making you a multimillionaire.
Does the food scientist now own part of your company because you used his flavor? Of course not. There’s no doubt your business benefited from his research. Without the flavor he created, your soda might never have existed. But the scientist has already been paid — and it’s worth remembering that he obtained that payment by force. The idea that he’s entitled to additional compensation because you successfully used his research is absurd. Contributing an input does not automatically confer ownership over everything built with that input. Likewise, the fact that government-funded research contributed to AI development does not automatically give the government ownership rights in AI companies.
In fact, the “government investment” argument fails even in a friendlier scenario that doesn’t involve coercion. Imagine that you voluntarily hire the food scientist to develop a new flavor. You agree to pay him $100,000 for his research, and he accepts. You then use the flavor he creates to launch a highly successful soda company.
Again, the food scientist does not own part of your company merely because he contributed to its success. He is entitled to whatever compensation the agreement specifies — nothing more and nothing less. If the contract grants him an equity stake, he has a claim to it. If it doesn’t, he doesn’t. That’s how market exchange generally works. Suppliers of inputs do not automatically acquire ownership of outputs. If the government wants to fund research in exchange for equity, it can try to negotiate those terms in advance. But that’s very different from claiming, after the fact, that successful companies partly belong to the government because some of the knowledge they relied upon emerged from government-funded research.
If merely contributing to a business entitled someone to partial ownership, every business would face endless competing claims from the countless people who helped make it possible. Imagine your local barista claiming partial ownership of the house you built ten years ago because the coffee she sold you helped fuel its construction.
This matters because property rights determine who gets to make decisions about how resources are used. A workable system of property rights requires clear rules for identifying who controls an asset. Without such rules, individuals and firms could never know whether they were free to use, improve, or exchange a resource. In market economies, ownership is generally determined by voluntary agreement, not by the mere fact of having played some causal role in a project’s success. The food scientist is paid as agreed; he does not acquire an open-ended claim on everything his work makes possible. Likewise, even if government-funded research contributed to the development of AI, it does not follow that governments thereby acquire ownership rights over AI companies. Causal contribution may explain how something came into existence, but it is not enough to establish a right to control it.
