On 9 May 2026, the the General Office of the Ministry of Industry and Information Technology issued the Notice Regarding Implementing the Pilot Scheme for Artificial Intelligence Science and Technology Ethics Reviews and Services was issued. The pilot explores best practices and refines mechanisms to build a comprehensive AI ethics review and service system, linking central, provincial, and municipal efforts through multi-stakeholder collaboration.
It will establish professional institutions, develop over 5 standards, build an ethics risk case database, and create technical tools and model applications to standardize services, guard against AI ethics risks, and support responsible innovation and high-quality industrial development.
4 key actions were planned out:
- Establishing local governance frameworks by refining provincial review regulations and municipal coordination mechanisms across departments.
- Building ethics committees and review centers to provide independent oversight and professional services.
- Validating review procedures through practical assessments, expert reviews for high-risk activities, and standardization of proven practices.
- And constructing a three-tier agile governance network with information sharing, risk reporting, and early warning capabilities.
Notably, Action 3 addressed standardization’s role as a central pillar in the pilot program. Participating cities will conduct practical ethics assessments of AI activities using existing relevant guidelines for Artificial Intelligence Ethics Risk Assessment and Generative AI Ethics as reference standards. Organizations lacking internal ethics committees is directed to local review centers for support.
For high-risk AI activities, provincial authorities will establish expert review protocols and conduct at least 3 expert reviews during the pilot, feeding insights into the dynamic updating of the high-risk activity list. The pilot mandates the conversion of proven review practices into formal technical standards: cities are tasked with organizing innovators to contribute to the development and validation of standards covering ethics risk assessment and committee establishment, while also piloting AI ethics management system certification, ensuring that mature operational experience is systematically codified into standardized norms.
The program runs from 1 June to 30 November 2026. Provincial authorities nominate one to two pilot cities and submit implementation plans by 20 May 2026. MIIT monitors progress monthly, tracks ethics reviews and expert assessments, coordinates risk information sharing, and resolves operational issues. Upon completion, cities submit summary reports for evaluation; high-performing regions receive recognition, promotion at major events, and preferential policy and funding support.
This pilot provides a structured, time-bound opportunity to observe how China is building institutional capacity for AI ethics governance at the operational level. The six-month cycle and explicit deliverables spanning standards, risk case databases, and technical tools indicate a methodical approach to testing governance mechanisms before wider application. As results emerge, the pilot’s outputs on review methodologies and multi-tier coordination structures will offer practical reference points for understanding the current operational baseline of Chinese AI ethics requirements. SESEC will continue monitoring the development of the pilot program and report potential regulatory changes and key standardization updates.
Source: https://www.miit.gov.cn/zwgk/zcwj/wjfb/tz/art/2026/art_353d0d2a06ab4cebbfd53f97aec4ae24.html
