China’s Spy Chief Warns U.S. AI Models Pose New National Security Risks

China’s top intelligence official, Chen Yixin, has issued a detailed warning about the security implications of rapidly advancing artificial intelligence. In an article published in China’s Cyberspace journal, Chen argued that AI is entering a new phase in which automated systems can significantly increase the speed and scale of cyber operations.

Chen specifically mentioned Anthropic’s Claude Mythos and OpenAI’s GPT-5.5-Cyber, saying that such systems demonstrate a major increase in the ability to discover software vulnerabilities, develop malware and potentially automate complex cyber operations. Importantly, his article did not allege that either model had actually been used in an attack against China.

The Chinese official also warned about AI-generated deepfake audio and video, synthetic images and text, and automated bot networks. He said these technologies could be used to spread political rumors and misleading information on a large scale. One concern is that convincing AI-generated material could make it increasingly difficult for people to distinguish authentic statements from fabricated ones.

Cybersecurity was another major concern. Chen argued that AI is lowering the technical and financial barriers to sophisticated cyberattacks by helping automate parts of the process. He described an emerging environment in which automated systems could be used on both sides of cyber conflicts, potentially creating what he characterized as an “AI versus AI” phase of cybersecurity.

The warning comes amid an increasingly intense debate between the United States and China over AI development and governance. While some U.S. technology leaders have been calling for stronger safety measures around advanced AI, Beijing has simultaneously criticized efforts to restrict China’s technological development. Chen’s comments therefore add another dimension to the international AI debate: governments are increasingly treating advanced AI not only as an economic technology but also as a matter of cybersecurity, information security and national strategy.