From July 8 to July 10, 2025, the Artificial Intelligence Subcommittee of the National Information Technology Standardization Technical Committee (SAC/TC28/SC42) convened its annual Artificial Intelligence Standards Week in Beijing. More than 600 representatives attended the meeting, including officials from relevant government departments, enterprises and public institutions, universities and research institutes, industry associations, and news media. The event commenced with two plenary meetings. One is from the National Artificial Intelligence Standardization General Group, while the other one is from SC42.
The National Artificial Intelligence Standardization General Group was established in 2018 under the State Council’s notice, Issuing the Development Plan on the New Generation of Artificial Intelligence (2017). It is responsible for the overarching coordination and top-level planning of artificial intelligence standardization, including:
- Formulating the planning, policies, and layout of the national AI standards system;
- Coordinating the attribution of technical content in standards to the relevant committees;
- Establishing a conversion mechanism between basic standards and industry application standards.
The Expert Advisory Group serves as a “think tank” for the Overall Working Group, providing strategic direction, policy recommendations, and assessments of development trends, and has played a pivotal role in shaping China’s AI standards system.
On behalf of the General Group, Mr. Yang Xudong, President of the China Electronics Standardization Institute (CESI) presented the ‘National Artificial Intelligence Standardization Work Report‘. The report systematically reviewed important phased achievements in China’s AI standardization efforts, introduced progress in areas including the standard system framework, development of key standards, standard verification and application, and international standardization cooperation. It also clarified future standardization work deployments in critical fields such as computing power benchmarks, large model evaluation benchmarks, large model selection and application guidelines, intelligent agent protocols, embodied intelligence, and AI safety governance.
According to the work report, China has published 27 national standards in AI, with 20 more in development, and 3 standard pre-research projects underway. In addition, 25 sector standards for AI have been published; 127 sector standards are in development, and 45 standard pre-research projects are ongoing.
Building on this foundation, the General Group will prioritize the standards development and relevant research in five key technical areas: AI computing power benchmark, large model evaluation system, guidelines for large model selection and application, embodied intelligence, and agent protocols.
The working meeting concluded with a comprehensive review of China’s AI standardization progress and the establishment of key development priorities.
The agenda then proceeded to the SC42 Plenary Meeting, where representatives from all 17 working groups presented status reports covering 43 published standards and 87 standards in active development, detailing their respective technical scopes and implementation timelines.
The subsequent sessions on 9 and 10 July were dedicated to working group discussions, with participation from approximately 2,000 industry experts and subcommittee members. Detailed reports for working group level discussions were made specifically by SESEC team and will be distributed to EU stakeholders as below.
