China says it will speed up a comprehensive AI law

An official of China’s National Development and Reform Commission (NDRC), the state planning agency, said on 3 November that China will accelerate basic, comprehensive legislation on artificial intelligence. Wang Ruomeng made the remark at a press conference in the NDRC’s series on the 15th Five-Year Plan period, and the NDRC published the Chinese-language transcript the same day.

Wang said there was now broad consensus in society that AI brings opportunities and risks together, and cited Xi Jinping’s call to keep AI safe, reliable and controllable. She set out four priorities. The first is a governance rule system: summing up China’s regulatory experience to push forward the comprehensive law, more legislation on shared factors such as data, computing power and algorithms and on key application scenarios, safety standards and ethical norms covering the whole AI life cycle, and periodic evaluation of whether rules still fit. The second is more adaptive regulation, with transition periods, compliance guidance and support services, and regulatory sandboxes for different sectors. The third is shared responsibility among government, industry bodies, companies and the public; for companies, Wang said the state would push developers and operators to disclose safety risks promptly and to report safety incidents and vulnerabilities. The fourth is social impact, covering employment monitoring, algorithmic discrimination, deepfakes, privacy, and protection of the elderly and minors, along with guidance on emotional dependence on AI.

The transcript contains no draft law, no timetable and no indication of which body would lead drafting. The remarks state a direction rather than create obligations. They add to a recent run of Chinese policy documents, including an expansion of the national AI safety framework to cover agents and proposed export controls on AI models, agents and computing. The reference to incident reporting also comes about six weeks after Washington and Beijing agreed to an AI incident channel, though the transcript does not mention the agreement.

On 9 November the Cyberspace Administration of China announced a single online filing system at beian.cac.gov.cn for generative AI, algorithm, deep synthesis and blockchain services. It replaces the separate algorithm-filing and blockchain-filing systems, which stopped operating the same day. The notice changes where filings are made; it does not say who must file.