China’s top intelligence official has issued a stark warning about the risks artificial intelligence (AI) poses to the Chinese Communist Party’s grip on power, outlining the technology’s potential to undermine political stability, national security, and military competitiveness.
Chen Yixin, head of China’s Ministry of State Security, detailed these concerns in an article published Sunday in the state-run magazine China Cyberspace. The commentary represents the most comprehensive public assessment yet by a senior Chinese official on the security challenges posed by AI and underscores Beijing’s intent to tighten party control over the technology.
Chen emphasized the need for increased government oversight to counteract multiple threats associated with AI. These include the potential for “hostile forces” to deploy AI-driven propaganda campaigns using deepfakes and other fabricated content to destabilize the party’s authority. Experts noted the fears that fabricated statements appearing to come from party leaders could sow confusion and unrest.
In addition, Chen warned that AI lowers the technical and financial barriers for cyberattacks on critical infrastructure, heightening risks to China’s information systems. The warning follows recent disclosures from U.S.-based researchers who demonstrated vulnerabilities in WeChat, China’s largest messaging platform, which could be exploited at scale.
The article also highlighted dangers from the use of foreign AI models by Chinese users, which might lead to significant data leaks, including sensitive information linked to the Chinese military and top technology firms. Chen named U.S. technology firms and their AI systems, such as models developed by Anthropic including Claude Mythos and GPT-5.5-Cyber, as sources of concern. The Chinese Communist Party views these foreign AI tools as a direct challenge to its data sovereignty and cybersecurity.
On the military front, Chen pointed to the evolving nature of warfare underpinned by AI, citing the U.S. military’s deployment of AI tools in recent conflicts as an example of the need for China to accelerate its own AI capabilities to maintain battlefield initiative.
Observers interpret Chen’s message as signaling that China approaches forthcoming U.S.-China discussions on AI safety — expected ahead of a meeting between President Xi Jinping and former U.S. President Donald Trump scheduled for September 24 in Washington — through an adversarial lens. According to Henry Gao, a law professor at Singapore Management University, Beijing views AI-related challenges primarily in terms of national security competition rather than as a shared global concern.
Chen’s article follows a series of previous warnings from various Chinese regulators about specific security risks linked to AI products, including concerns over Anthropic’s coding tool Claude Code. Chinese authorities have also cautioned domestic users about the vulnerabilities inherent in AI assistants.
At the same time, China's Foreign Ministry has criticized what it describes as exaggerated portrayals of AI threats by U.S. companies and officials. The ministry rejected narratives suggesting that China’s AI industry poses a direct threat to the United States, accusing such rhetoric of worsening tensions and complicating international regulatory cooperation.
This exchange came amid calls from some Western AI executives, notably Anthropic CEO Dario Amodei, for Washington to impose restrictions on the sale of advanced AI chips to China. Amodei argued that keeping technological advantages is vital for democratic countries to maintain a lead over China’s AI development.
Chen’s commentary underscores Beijing’s dual approach: recognizing AI as a strategic asset essential for national development and military strength, while emphasizing the imperative of securing the party’s control and protecting China’s vital information infrastructure in an increasingly competitive and contested global technological landscape.