As AI cyber risks grow, Chinese chip expert urges shared ethics standards with US rivals

Chinese AI chip pioneer Chen Yunji has called for the establishment of universal ethical guidelines for both Chinese and American AI models to address growing cyber risks, according to a report from the World Laureates Forum in Shanghai. Speaking at an event on Friday, Chen highlighted the rise of high-profile cyberattacks that have targeted critical infrastructure and government servers, urging a "universal ethics checker" for all leading AI models, including those developed by Chinese companies DeepSeek and ByteDance, and American start-up OpenAI.

While acknowledging the potential for AI systems to autonomously improve themselves (recursive self-improvement), Chen expressed faith in leveraging AI for safeguards at each development stage. He emphasized that achieving consensus on these ethical standards is crucial to prevent harm to humanity. The Chinese government has recently proposed export controls on advanced AI systems and additional restrictions for those exceeding cyber-hacking thresholds.

Experts like Martin Hellman, a Turing Award recipient, stressed the need for cooperation between China and the US to ensure the responsible governance of AI, warning that current economic and military motivations could lead to catastrophic outcomes.