It’s a familiar refrain in tech boardrooms and government corridors: China is aggressively challenging America to an arms race in artificial intelligence. Whoever wins will control the geopolitical landscape — and global economy — for generations.
For years, Pentagon officials and tech luminaries like Eric Schmidt, Peter Thiel, and Alex Karp have parroted this doomsday scenario.
But the narrative doesn’t hold water.
There’s compelling evidence that China’s AI capabilities have been overestimated. While Chinese technologies have advanced rapidly, they aren’t an imminent national security threat. Consider this alternative perspective from retired U.S. Air Force Lieutenant General Jack Shanahan: “It feels at times like we are dangerously close to making the same kind of erroneous ‘bomber-missile gap’ assessment with AI that we did with the Soviet Union in the early 1960s.”
The Cold War-era bomber–missile gap was illusory, but it sparked a nuclear arms race with devastating political, environmental and psychological consequences. It was also extraordinarily expensive, costing American taxpayers $5.8 trillion, according to one estimate. That’s more than $11 trillion today.
A rational geopolitical analysis should take China’s challenges into account. The country’s economy has been decelerating for 20 years, with no end in sight. A demographic decline will cut China’s population in half by century’s end. A brain drain is pulling talented Chinese researchers to Australia, Canada and the European Union. As Taiwan’s outgoing president Tsai Ing-wen said last fall, China is too overwhelmed with domestic problems to invade Taiwan.
Why, then, has the idea of a cataclysmic battle between “AI superpowers” taken hold? The answer is deceptively simple: For some, war is profitable. Preparing for algorithmic warfare is even more so.
The China AI threat narrative boosts the Pentagon’s demand for high-tech weapon, surveillance and logistical systems, serving to justify and accelerate U.S. defense tech spending, while bending government AI research toward military rather than civilian uses. Tech companies are lining up to claim their share of the Defense Department’s $886 billion annual budget — not only giants like Amazon, Microsoft, Google, and Palantir, but also hundreds of Silicon Valley defense tech startups financed by venture capital.
These firms are also doing business with foreign militaries, including Google, whose $1.2 billion contract with the Israeli defense ministry has sparked protests among employees who are appalled by the company’s support for the war in Gaza. Google fired 28 employees earlier this week for protesting.
These changes are transforming the nature of the military-industrial complex. In 1961, President Eisenhower warned Americans that unfettered defense spending threatened democracy by giving defense firms “unwarranted influence” over Congress.
Today, this is still true. Dozens of former Pentagon officials are now working for venture capital firms. The “revolving door” between the Pentagon and private industry still spins, but military officials are as likely to gravitate toward Silicon Valley investment firms than they are to Lockheed Martin, Raytheon or other “traditional” defense companies.
This will have dire consequences for future wars. High-risk defense tech startups prioritize rapid growth, profitable business models, aggressive marketing campaigns and accelerated “hype cycles.” Meanwhile, corporate executives make extraordinary but unverifiable claims about their products.
While such practices may be acceptable for consumer goods, there’s much more at stake when Silicon Valley’s startup model is applied to military products especially AI-enabled weapon and surveillance systems that make unpredictable decisions, or fail when the operating environment changes. Deploying inadequately tested technologies will result in innocent people and American troops dying.
If the pace of developing and adopting military AI weapon and surveillance systems continues accelerating, the end result will be an expensive arsenal of flawed, unreliable and dangerous technologies that don’t work as advertised.
Roberto J. González is a professor in the anthropology department at San José State University. His most recent book is “War Virtually: The Quest to Automate Conflict, Militarize Data, and Predict the Future.”