Full text
- This is an unofficial copy. The document has been archived and reformatted in plaintext for AGORA. Footnotes, tables, and similar material may be omitted. For the official text, visit the original source.
- This text may be out of date. According to the latest data in AGORA, this document has been proposed, but is not yet enacted or otherwise finalized. This text was collected 2024-10-29 and may have been revised in the meantime. Visit the official source for authoritative text.
5. Key recommendations for AI standardization work
Based on the current state of AI technology and industry development, together with the
progress of standardization work and the establishment of the standards system, we propose key
work recommendations for China’s AI standardization.
(1) Improve the working mechanism and help industry to achieve healthy and
sustainable development
AI is deeply integrated into the real economy, and the standardization of AI is exhibiting a
trend of collaborative participation in multiple fields. First, through the National Information
Technology Standardization Technical Committee’s AI Technical Subcommittee (SAC/TC
28/SC 42), the standardization of new generation information technology fields such as AI, the
Internet of Things (IoT), cloud computing, and big data should be coordinated and promoted to
strengthen the application of AI in vertical fields. Second, the role of the national AI
standardization working group platform should be given full play to coordinate the development
of AI standards in the real economy and related fields, and to absorb the energy of government,
industry, academia, research institutes, and users to jointly promote AI standardization.
(2) Develop key standards and improve the AI standards system
The development of key AI standards must be promoted based on the principles of
“advantages first, maturity first, lead with basics, and be application-driven.” The first is to
consolidate the industrial foundation with a focus on the development of standards such as
performance benchmarks, hardware virtual interfaces, and development framework compatibility
specifications. The second is to support industrial applications and develop a number of
industrial application standards around areas with a certain industrial scale, such as intelligent
speech and computer vision. The third is to promote integrated development with the goal of
supporting the deep integration of AI into the real economy and to develop a batch of industry
application standards for applications such as smart transportation, smart healthcare, and smart
classrooms.
(3) Carry out pilot demonstrations to improve industrial service capabilities
China’s AI standardization work is progressing steadily, and relevant standards are being
successively released. Governments and relevant departments at all levels must play a guiding
role to strengthen the promotion of AI standardization. The first is to carry out propaganda and
promotion activities around key standards such as machine learning models and machine
translation capability evaluations, combining the application of standards with the development
of the AI industry. The second is to promote the construction of a standard compliance
assessment system, focusing on selecting national AI innovation application pilot zones and
national new generation AI innovation development test zones to carry out standard application
demonstrations. The third is to build an AI standardization public technology service platform,
improve the service capability of standardization for industry, and support basic and general
purpose technology research and development, third-party evaluations, and other such work.
(4) Create de facto standards (事实标准) and enhance industry competitiveness
A de facto standard is the assassin's mace (杀手锏) of industry competition. By
summarizing the outreach experience of information technology standards such as DCMM,
ITSS, and CMMI, a cultivation model should be explored for China’s de facto AI standards. By
combining major works and major projects around the open-source community and new
infrastructure construction, and through consulting on training, testing and evaluation, product
certification, and other means, standards should be promoted with large-scale promotion and
applications to create a batch of de facto AI standards.
(5) Strengthen international exchanges and promote the global promotion of
Chinese wisdom
The first is to actively connect with international standards organizations such as ISO/IEC
JTC 1 and continue to strengthen participation in AI international standards. The second is to
rely on the National Artificial Intelligence Standardization General Working Group and the
National Information Technology Standardization Technical Committee Artificial Intelligence
Technical Subcommittee to organize the participation of domestic standardization institutions
and enterprises in the formulation and revision of international standards and to contribute
Chinese wisdom to international standard proposals. The third is to support domestic experts in
assuming the positions of secretary and convener in the International Organization for
Standardization to enhance the right to speak internationally. The fourth is to actively host
international conferences and forums related to AI, deepen bilateral and multilateral cooperation
in the field of AI, and achieve mutual benefits.