The Shared Threat

On May 6, the New York Times published an article titled “The Shared Enemy of China and the U.S.” by Thomas Friedman, highlighting a new common threat facing both nations.

China and the U.S. are now confronted with a novel risk of disorder that could undermine global stability and harm their interests if they fail to cooperate while competing. Only through joint action can they effectively address these challenges.

A World of Integration

Given the remarkable capabilities of the latest artificial intelligence (AI) models in cyberattacks, China and the U.S. should collaborate to establish protective mechanisms against the malicious use of AI.

Since President Nixon’s visit to China in 1972, two paradigm shifts have reshaped the world. The first is the emergence of new asymmetric AI tools that empower small malicious actors, including terrorists, anarchists, criminals, political groups, and small nations. The second shift relates to globalization. Nixon’s visit initiated a process that transformed the world from isolation to close connection and deep interweaving. In the early 21st century, China’s accession to the World Trade Organization and the proliferation of the internet allowed more regions and individuals to compete, connect, and collaborate at lower costs and in more ways. This led to my writing “The World is Flat” in 2005.

However, technological change means that each significant advancement occurs faster than the last, building on the tools released by previous eras. As noted by Dov Seidman, founder of the LRN Corporation, the world has shifted from being interconnected to interdependent, or from “flat” to “integrated.”

In a flat world, one could choose to unplug; in an integrated world, no one can escape. Today, our fates are intertwined, and we are inextricably linked.

This is not only due to the internet, smartphones, fiber optics, satellites, and wireless communication technologies connecting us in unprecedented ways, but also because a series of global challenges have intertwined our fates like never before. These challenges are vast and borderless, and no nation, regardless of its strength, can address or evade them alone.

The Threat of Proxy AI

We know what these challenges are: mitigating climate change, preventing the proliferation of nuclear and biological weapons, managing global migration, controlling pandemics, ensuring the smooth operation of global supply chains, and, most importantly, regulating the new types of AI we have created.

While we can manage some global issues through limited cooperation, we no longer have time to delay in addressing the cyberattack capabilities of AI. There is no room for procrastination; we have reached a point of no return.

Now, guess who has entered the fray? A new set of potentially dangerous actors, no longer limited to nation-states, can threaten both China and the U.S.

This refers to the recently disclosed proxy AI systems by Anthropic and the OpenAI Research Center. Such systems can provide small-scale cyber attackers with tools to disrupt the economies of China, the U.S., and any country globally at a very low cost and with minimal technical expertise.

Craig Mundie, former head of research and strategy at Microsoft and my mentor in researching new AI threats, stated, “The two countries must lead in controlling their proliferation and building defensive systems. They need to protect themselves while preventing technology leaks that could harm others.”

Mundie believes, “This should become a powerful motivation for cooperation between the two countries, even if limited to this narrow issue, as it now poses an imminent threat to both.”

Collaborating with Tech Companies

This is not an unrealistic demand.

China and the U.S. should collaborate with what I call the “New I7”—seven companies (Anthropic, Google/Alphabet, OpenAI, Metaverse platforms, Alibaba, DeepMind, and ByteDance)—to harness the value of new AI models while effectively mitigating their worst impacts. Governments alone cannot solve this issue; businesses are equally unable to tackle it alone.

Due to the impact of the Iran War, this dynamic has not received the attention it deserves. Reports indicate that President Trump is considering regulating AI models before they are made public. We must recognize that we are entering a new era where the power held by private enterprises is comparable to nuclear fission, with impacts radiating across various fields.

“Like nuclear fission, this technology can be used for power generation or to create bombs,” Mundie said. Proxy AI is similar; it can generate immense benefits or create weapons with asymmetric destructive power.

Although many global leaders have yet to fully grasp this, humanity has entered a new phase in history: we must govern, innovate, collaborate, and coexist on a planetary scale to achieve prosperity. We must either build complex and adaptive coalitions to achieve this goal or face collective decline.

Our fates are now intertwined.