Chen, who heads the Ministry of State Security, set out the risks in an article published on Sunday by China Cyberspace magazine, arguing that Beijing must build a stronger system for monitoring, regulating and responding to dangers created by rapidly advancing AI.
His intervention came as Anthropic chief executive Dario Amodei urged leading AI developers to slow the pace of frontier-model development so that safeguards can keep up with increasingly capable systems. OpenAI chief executive Sam Altman and xAI founder Elon Musk have also backed calls for greater restraint, sharpening an international debate over whether commercial and geopolitical competition is outpacing safety controls.
Chen said hostile actors could exploit deepfakes, AI-generated text and images and automated online networks to manufacture political rumours, spread harmful material and inflame social divisions. Such activity, he argued, could threaten China’s political, institutional and ideological security.
A second risk involved cyber warfare. Chen said advanced models were increasing the efficiency of vulnerability discovery, malware development and weaponisation, lowering the technical threshold and cost of attacks while pushing cyber conflict towards greater automation. He warned that China’s critical information infrastructure could face serious exposure as offensive AI tools become more capable.
He also highlighted the danger of large-scale data leakage. Some users in China, Chen said, lacked sufficient security awareness and could expose sensitive information by using overseas AI products or sending data abroad. Beijing has tightened scrutiny of cross-border data flows while encouraging domestic development of models and computing infrastructure.
Chen identified technology concentration as another danger, accusing unnamed countries of using export controls, entity lists, technical restrictions and closed ecosystems to preserve advantages in advanced computing and AI. China has repeatedly criticised United States restrictions on exports of high-end semiconductors and chipmaking equipment, while Washington says the controls are intended to limit military and national-security applications.
The minister’s fifth concern was the structural impact of AI on social governance, while the sixth was its capacity to alter warfare. He said algorithmic systems were increasingly being used for intelligence fusion, decision support, target identification, operational assistance and information operations, making AI a more central element of military competition.
Chen called for an AI security regulatory platform, stronger risk monitoring and early-warning systems, regular security guidance and expanded research into emerging technological threats. He also urged greater investment in basic AI research, computing chips, development frameworks, application software and national data infrastructure.
Chen’s article also called for stronger legal safeguards, public education and international cooperation, while urging faster construction of national and local data pools and computing networks designed to support domestic AI development securely nationwide.
The warning reflects a broader shift in Beijing’s public discussion of AI. China has promoted the technology as a strategic driver of economic growth while increasingly stressing that advanced systems must remain safe, reliable and controllable. A statement issued after the World Artificial Intelligence Conference in July called for safeguards around frontier models and warned that AI could pose threats to critical infrastructure in finance, electricity, telecommunications and transport.
Amodei’s appeal takes a different approach but overlaps with Chen’s emphasis on cyber and systemic risks. The Anthropic chief proposed giving independent evaluators continuing access to advanced systems, coordinating safety practices among leading developers and pursuing international arrangements intended to reduce incentives for an uncontrolled race.
Anthropic’s September threat-intelligence report said its Claude models had been misused or targeted for misuse in cyber operations, surveillance, influence activity, fraud, weapons-related research and other harmful activity. The company said it disrupted accounts linked to several operations, including China-based actors using AI to automate surveillance and intelligence tasks.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
