Investing.com -- What began as an economic rivalry over microchips has taken a chillingly dark turn, as Washington analysts and Chinese state commentators openly weigh a scenario once confined to science fiction: direct military strikes on artificial intelligence infrastructure.

The escalatory rhetoric reached a sharp peak after Hu Xijin, former editor-in-chief of the state-backed Global Times, directly threatened the U.S. mainland in response to American policy proposals. Writing on X (formerly Twitter), Hu warned: "If the U.S. truly dares to launch a military strike against China’s data centers, I believe China’s retaliation would not be limited to U.S. targets in the Asia-Pacific region; instead, it would directly target the U.S. mainland with missile strikes."

Hu’s threat highlights the high-stakes fallout surrounding a new report titled "Superpowers and AGI" by Jacob Stokes, deputy director of the Indo-Pacific Security Program at the Centre for a New American Security (CNAS). Stokes, a former Obama administration national security staffer, called Friday for the United States to prepare contingency plans for military strikes on Chinese data centers, state-backed espionage, and cyber operations to prevent Beijing from achieving artificial general intelligence (AGI) first.

The proposals land one day after OpenAI president Greg Brockman declared "Welcome to the AGI era" at the launch of GPT-6 Astra on September 3, a milestone that sharpens the urgency of every scenario Stokes describes.

Stokes framed the challenge as analogous to the nuclear arms race, arguing that policymakers must "work backwards based on the unique nature of the technology, in the same way that in a past era, policymakers would learn about nuclear weapons" to derive policy from science. He told an online event that the Department of Defense and National Security Agency should begin assessing what intelligence would be required to justify extreme measures, including preventive strikes before China develops AGI.

Neither the Department of Defense, the National Security Agency, nor Beijing has officially responded to Stokes’s proposals.

The U.S. currently holds a roughly seven-month lead over China in frontier AI model quality, according to Epoch AI. But Stokes warned that China could achieve a "surprise technological breakthrough," pointing to advances in robotics and other forms of embodied AI as a potential pathway.

Stokes also raised the option of reverse-engineering or stealing Chinese AGI technology, arguing it would be justified "legally and normatively" given China’s alleged history of intellectual-property theft. He acknowledged, however, that the U.S. intelligence apparatus has "not been organized for espionage to gather technological secrets for decades," making structural reform a prerequisite.

Not everyone shares Stokes’s calculus. William Hartung, an arms analyst at the Project on Government Oversight, told the South China Morning Post that strikes on Chinese data centers would "risk starting a shooting war between two nuclear-armed powers that could have tragic consequences for all concerned," calling the idea "the height of recklessness."

The debate unfolds against a delicate diplomatic backdrop. Trump and Xi Jinping are scheduled to meet in Washington on September 24, with Treasury Secretary Scott Bessent confirming that AI governance is on the agenda, as CNBC reported on September 2. Stokes’s own assessment of that summit is muted: the "best-case scenario" is that both sides find the talks "constructive" rather than producing any governance breakthrough.