Former US Secretary of State Henry Kissinger, who passed away in November 2023, delivered a video address at the World Economic Forum in Davos, Switzerland, in January of that year. At age 100, in a still-clear voice, he warned that the risk of armed conflict between the United States and China — both nuclear-armed powers developing artificial intelligence (AI) capabilities — needed to be examined in meticulous detail. He noted that the two countries had narrowly avoided a path toward military confrontation thanks to a meeting between US President Joe Biden and Chinese President Xi Jinping roughly two months before the lecture. That summit had been dramatically arranged through his own strenuous mediation, shuttling between Beijing and Washington, D.C. At the time, Chinese fighter jets had repeatedly crossed Taiwan's Air Defense Identification Zone (ADIZ) to threaten the Taiwanese mainland, while the United States countered by transiting the Taiwan Strait through "freedom of navigation" operations.
In that address, Kissinger assessed that rapidly advancing AI technology was penetrating deep into the military and administrative networks of governments, beyond daily life, significantly raising the possibility of full-scale war between major powers. This suggested that the AI revolution would pose a graver threat to humanity's survival than any previous scientific revolution.
Kissinger, who won the Nobel Peace Prize in 1973, was renowned not only for his work in diplomacy and security but also for his exceptional perspective on the currents of future advanced industries. His 2021 book, "The Age of AI: And Our Human Future," vividly depicted a scenario of full-scale war between the United States and China that Koreans — facing nuclear threats from North Korea — should examine closely. He wrote that if North Korea decided to strike the US mainland, the United States would immediately detect it and launch interceptor missiles. But at that very moment, China would also detect the action and mount a counterattack against US missiles, according to his analysis. Remarkably, he said the North Korean provocation and the onset of full-scale US-China war would unfold in just 0.003 seconds. Kissinger argued that AI would trigger future wars in instants too brief for humans to even perceive. This, he said, is what future warfare looks like.
Kissinger's chilling prophecy has grown more likely to materialize with the emergence of "monster AI hackers" capable of designing and executing cyberattacks on their own. Earlier this month, US AI company Anthropic unveiled its latest model, "Mythos," poised to upend the global cybersecurity landscape. After completing testing of the next-generation AI, code-named "Capybara," Anthropic withheld the model from public release. Although its performance set new records, the cybersecurity risks were so vast as to defy imagination, and the company locked the model down on its own out of concern over the potential fallout.
Just how catastrophic are the risks that prompted the ban on public release? Mythos discovered bugs within minutes that security experts had failed to find for decades. More astonishingly, Mythos did not stop at identifying flaws — it also autonomously generated code capable of paralyzing systems or launching hacking attacks. Anthropic imposed a forced seal on Mythos out of a sense of crisis: if the AI model were to fall into civilian hands and penetrate the security networks or nuclear attack programs of a particular country, it could quite literally become a terrifying disaster for humanity. Yet this is no cause for relief. Given the AI race among Big Tech firms, the possibility that other major AI companies will build models similar to Mythos cannot be ruled out.
The full-scale war between major powers in the AI era that Kissinger prophesied — even before ChatGPT made its full-fledged debut — and the risk of armed conflict between South and North Korea caught in between have now become an imminent reality. North Korea, which clings to hacking as a means of regime survival, is unlikely to take AI hackers lightly. The prospect of a powerful AI hacker model falling into the hands of extreme terrorists is terrifying to contemplate. Moreover, if North Korea were to obtain a "digital nuclear weapon" like Mythos, the chilling US-China war scenario Kissinger warned of could overlap with the Korean Peninsula's future.
Against this backdrop, the recent discord between Seoul and Washington over the controversy surrounding the leak of North Korean nuclear intelligence is worrying. The fallout from the intelligence leak has already escalated to the point where the United States has suspended the sharing of information related to South Korea and North Korea. The mistrust between the two allies must be resolved as quickly as possible so that the friction caused by the leak controversy does not lead to damage to the South Korea-US alliance — damage that could shake the future of the Korean Peninsula. Kissinger said that "every civilized society must defend itself, either through its own strength or through alliances." That is why South Korea, perpetually under North Korean nuclear threat, must once again take deeply to heart Kissinger's advice emphasizing the power of alliances.
