The Pulse
China Plans AI Rules Covering Data, Computing and Deployment
China’s National Development and Reform Commission says it will accelerate AI legislation and develop safety and ethics standards covering the technology’s full life cycle. The plan names data, computing power, algorithms and key applicatio
AI.info Team · Written with AI · fact-checked ·
“We need to establish safety standards and ethical guidelines covering the entire life cycle of artificial intelligence.”
Wang Ruomeng, deputy director of the National Development and Reform Commission’s Department of Innovation and High-Tech Development
China plans to accelerate foundational AI legislation and build safety and ethics standards that reach across the technology’s full life cycle, according to remarks Wang made at a news conference on November 3. The proposed rules would address shared inputs including data, computing power and algorithms, as well as legislation for key application areas.
Wang said the work would take place during China’s 15th Five-Year Plan period. Authorities also intend to assess how the rules work at regular intervals and adjust provisions that no longer fit as AI technology changes, Xinhua Finance reported.
Data and computing join application rules
The announcement frames AI governance as more than oversight of services after they reach users. By naming data, compute and algorithms alongside applications, it points to rules that could cover multiple stages of model development and use. Wang did not set out specific obligations for developers or operators in the remarks carried by Xinhua Finance.
The plan also calls for summarizing China’s existing AI-governance experience and clarifying compliance boundaries. The report did not include proposed statutory language, a legislative schedule, enforcement provisions or penalties. It describes work to come, not a newly enacted law.
Existing measures cover narrower parts of the chain
China has already issued rules addressing specific parts of AI development and deployment. Guidelines released in May by the Cyberspace Administration of China, the National Development and Reform Commission and the Ministry of Industry and Information Technology address AI agents, setting out safety requirements across development, deployment, application and maintenance. They also call for risk controls such as testing and filing requirements in sensitive sectors, with lighter measures for some lower-risk uses.
In March, ten government bodies issued an interim framework for ethics review of AI-related scientific and technical activities. The measures cover review procedures and ethical principles for research and development. The November announcement reaches beyond those defined areas by proposing basic, cross-cutting AI legislation and standards spanning the full life cycle.
Periodic review is part of the proposal
Wang’s call for regular evaluation acknowledges a practical problem for any rulebook covering a technology whose capabilities and uses can shift: requirements written for one stage of development may become mismatched to later systems. The announcement does not specify how often reviews would happen, which agencies would conduct them or how companies and researchers would take part.
For now, the firmest details are the categories Beijing wants to address—data, computing power, algorithms and key applications—and its intention to revisit rules as technology advances. The next concrete step would be draft legislation or standards setting out who must comply and what those requirements entail.