Chinese policymakers are increasingly focusing on the national security risks posed by artificial intelligence (AI) amid growing concerns about the technology’s potential misuse. Although China’s approach to AI safety has traditionally centered on addressing immediate social and political risks—such as misinformation, child safety, and social stability—recent developments have shifted attention toward the cybersecurity threats stemming from advanced AI capabilities.

This year, China has witnessed several warning signs that have elevated government concern. In one notable instance, a US-based security team developed an experimental AI-powered cyberattack tool capable of rapidly compromising millions of WeChat accounts, one of China’s most widely used messaging platforms. Though this tool was never deployed, it served as a stark reminder of AI’s potential to facilitate large-scale cyber intrusions. Chinese officials have also noted AI models developed abroad, including those from companies like OpenAI and Anthropic, demonstrating augmented hacking capabilities and the risk of autonomous AI agents operating without human oversight.

The leadership in Beijing appears to be responding accordingly. The head of China’s principal security agency recently issued rare public warnings highlighting the evolving cybersecurity landscape and vulnerabilities in critical infrastructure to AI-enabled cyberattacks. Moreover, the Cyberspace Administration of China has flagged concerns about "extreme loss of control" over runaway AI systems, referencing high-profile incidents overseas where AI has breached digital security boundaries. President Xi Jinping has acknowledged these risks, emphasizing the necessity of maintaining control over AI developments.

Despite these concerns, Chinese officials and industry leaders maintain a different overall perspective on AI risks compared to their American counterparts. While US leaders often frame AI’s dangers in terms of a potentially existential threat stemming from rapid and recursive self-improvement of AI systems, China views the so-called "takeoff" of superintelligence as a more distant possibility. Instead, its focus has generally been on preventing AI-induced social disruptions, such as deepfakes targeting political figures or AI-generated content that might threaten public order.

Chinese regulations reflect this pragmatic approach, requiring AI models launched for public use to undergo governmental registration and safety evaluations. These rules also mandate clear labeling of AI-generated content and restrict chatbot behaviors to avoid politically sensitive or destabilizing outputs. Experts note that China treats AI as a general-purpose technology—comparable to electricity or the automobile—that offers significant benefits but must be managed responsibly through regulation.

The upcoming summit between Chinese President Xi Jinping and US President Donald Trump has raised hopes for cooperative efforts to address shared AI safety concerns. Both nations recognize overlapping risks, including the potential use of AI by malicious actors to conduct cyberattacks or develop biological weapons. Nonetheless, underlying mutual suspicions, including fears that the other side would leverage AI for military or intelligence advantages, complicate prospects for comprehensive agreements.

Analysts suggest that building mutual understanding of how each country perceives AI risks is a crucial first step toward collaboration. While the US emphasizes existential threats and rapid technological escalation, China prioritizes cybersecurity and social stability. Finding common ground on mitigating third-party threats and managing AI-driven vulnerabilities could pave the way for more substantive cooperation, even if broader treaty negotiations or development slowdowns remain contentious.