The United States is no stranger to technological arms races, having spent much of the Cold War in a two-pronged one-upmanship effort against the Soviet Union to build bigger rockets and land those on the moon. Its rivalry now is with China, and the latest battleground is artificial intelligence.
The United States is no stranger to technological arms races, having spent much of the Cold War in a two-pronged one-upmanship effort against the Soviet Union to build bigger rockets and land those on the moon. Its rivalry now is with China, and the latest battleground is artificial intelligence.
China has long sought to dominate the AI landscape, laying out a plan to become a “global leader” in the sector by 2030 and pledging billions of state dollars for research and development. U.S. breakthroughs have been more organic, illustrated most recently by the rapid global uptake of chatbots made by American companies, such as Google, Microsoft, and OpenAI, with Chinese counterparts largely playing catch-up. Experts caution, however, that applying the “arms race” framework to the development of AI doesn’t capture the global dynamics around the technology.
One fundamental distinction is the private sector’s role at the forefront of developing new AI capabilities: There were no private-built rockets in the 1960s. During the Cold War, key technological advances in the nuclear and space sectors were “characterized by a high barrier to entry and a near-monopoly by the state,” said Ryan Nabil, director of technology policy at the advocacy group National Taxpayers Union, whose research focuses on AI governance. “AI is characterized by lower barriers to entry, democratized access, and the preponderance of the private sector in driving innovation.”
This also isn’t a two-horse race. AI development is picking up pace around the world, and the technology’s underpinning of everything from email to missile systems raises the stakes significantly. A more useful historical analogy for the current state of AI development may be the revolution that occurred centuries before the space race.
“I think the Industrial Revolution in many ways is a more helpful paradigm,” said Paul Scharre, vice president at the Center for a New American Security. “It was this combination of technologies that were mutually supporting, much like we’re seeing today with AI and data, the internet of things, mobile devices, wired and wireless connectivity, this whole panoply of technologies.”
The Industrial Revolution, like today’s, also left an easy-to-understand scorecard. “You could see clear winners and losers in the 1800s depending on how rapidly countries industrialized,” he said.
So who’s winning the AI race? Multiple rankings in recent years that consider metrics such as investment, talent base, and research—including the number of patents and publications—have the United States leading the field, with China close behind and gaining rapidly, followed by countries such as the United Kingdom, India, and Canada. But that doesn’t paint the whole picture. AI isn’t just patents but processes.
“As a general observation, China seems particularly strong relative to the U.S. in application of AI as opposed to development of underlying AI technologies,” Nabil said. China leads in areas such as facial recognition, he noted, though that may be as much a function of Beijing’s governing style as any particular savvy with AI. Rather than picking winners and losers, U.S. companies could pick a lane.
“These instances raise the question of whether dominance in each AI subdiscipline is critical or even desirable. A better approach would be to identify subdisciplines within AI where the U.S. should maintain its global competitive edge and calibrate U.S. policy accordingly,” he added.
The heavy focus on research in evaluating a country’s position in the AI landscape—a metric that increasingly favors China—also often misses key context, said Emily Weinstein, a research fellow at Georgetown University’s Center for Security and Emerging Technology. “You can write as many papers as you want; that doesn’t necessarily mean that you’re able to commercialize that specific technology.”
Regardless, the United States and China’s application of a Cold War arms race dynamic to artificial intelligence reflects their desire to write the rules of the new game. The Biden administration’s efforts to ban TikTok and kneecap China’s semiconductor industry are in part aimed at slowing Chinese AI advances, and Washington’s increased focus on shaping AI policy mirrors what Beijing has already been doing for years.
“I’m not sure our goal should be to ‘win’ the competition as much as to continually shape it in ways that are conducive to our interests and, relatedly, our values,” said Jessica Brandt, policy director for the Artificial Intelligence and Emerging Technology Initiative at the Brookings Institution.
Just as with the AI chatbots that have taken the world by storm, it’s important to ask the right questions.
“What can the United States do to promote AI innovation? Such targeted questions can be more helpful than overly broad questions, such as ‘Is China winning,’ that are difficult to answer precisely,” Nabil said, adding that Washington needs a regulatory approach that gives U.S. companies latitude to innovate without telling them exactly how to do it.
Another big paradigm shift in how AI develops in the coming years is the broader U.S. effort to decouple from China. Chinese companies and scientists have long worked with their American counterparts on artificial intelligence, with firms such as Baidu setting up AI labs in Silicon Valley in the mid-2010s and researchers collaborating at U.S. universities. But cooperation has given way to antagonism.
“The U.S. and Chinese AI ecosystems are deeply intertwined, and we’re seeing both governments take steps to increasingly pull them apart,” Scharre said. While China has always closed off its internet ecosystem to the West, “it’s only recently that we’re seeing U.S. policymakers do the same.”
