In Silicon Valley, the deepest nightmare of scientists is that artificial intelligence (AI) develops too advanced and eventually gets out of control and launches a war or nuclear attack. But from Beijing's perspective, the response is extremely simple: pull the plug, cut off the Internet, and arrest those responsible.
The New York Times reported on the 27th that when top U.S. technology leaders warned that AI would pose a catastrophic threat to humanity and called for cooperation between the United States and China, this "AI doomsday theory" to many Chinese people is not only distant but also has a strong Western flavor. Under the Chinese system, the ruling Communist Party of China has strong physical control over society and believes that it can easily control any new technology that may destabilize the Internet or the epidemic in the same way it has controlled the Internet or epidemics in the past.
Please read on...
The report pointed out that China’s technology industry has a rather lukewarm response to “existential risks.” Many believe that AI is still in its early stages of development and the focus should be on innovation and opportunity. In addition, the Western narrative that sees AI as a "Frankenstein" that challenges the Creator and is punished does not exist in Chinese culture, which emphasizes the "unity of man and nature." China prefers to view technology as a tool for coexistence.
More realistically, many Chinese practitioners interpret the U.S.’s doomsday warnings as “fear marketing” to suppress Chinese AI companies. Anthropic CEO Dario Amodei recently called on democratic countries to maintain technological leadership and restrict China's access to advanced chips. Such hawkish remarks triggered a strong backlash from Chinese public opinion. Some people in the Chinese AI circle have even countered that if the United States monopolizes advanced AI, the threat would be no less than if Hitler obtained the atomic bomb first. Zhou Hongyi, founder of China’s information security giant “Qihoo 360”, also emphasized that the way to deal with risks is to “defeat AI with AI” rather than slow down the pace of development.
Despite their lack of interest in doomsday theories, Chinese authorities have not ignored the risk of runaway AI. With the advent of powerful AI models, Beijing has updated its "Artificial Intelligence Security Governance Framework" and begun drafting regulations to prevent AI from being exploited by hackers. China also implements the world's strictest industry regulations, requiring AI-generated content to be clearly labeled and models to be subject to official approval before release.
However, compared with the United States, Chinese AI companies still lack transparency in security protection. Chinese regulations do not mandate safety testing for catastrophic risks such as "making biological and chemical weapons." According to statistics from security consulting firm Concordia, in the past year, only half of China's top 10 leading AI companies have published security assessment results, and the details and frequency of disclosures are far less than those of their American counterparts.
In the context of a lack of mutual trust and geopolitical competition, although Chinese President Xi Jinping and U.S. President Trump agreed during their meeting in Washington that AI should be "managed for good," actual cooperation is extremely difficult to implement.
Sun Yuanyuan, founder of the "China AI Security Open Community" in Shanghai, pointed out that the United States' repressive remarks have made it more difficult for people in China who really care about AI security to promote relevant initiatives. Zhao Minghao, deputy director of the Center for American Studies at Fudan University, pointed out the essence of the U.S.-China standoff: "If the United States says they are slowing down, China will not believe it; similarly, if China says it is slowing down, the United States will never believe it."
