AI safety reads as a lifecycle security problem: corpus, model, infrastructure, and application security with assessment requirements attached, and a governance framework that links risk categories to management measures.
Standards BodyChina
National Cybersecurity Standardization Technical Committee
全国网络安全标准化技术委员会
Official siteThe committee where China's AI security expectations become testable documents: its guides pair with binding rules, its framework names the risk categories, and its new WG9 is where AI security standards are being drafted.
Importance
Assessed per aspect, band 0 to 3. There is no composite score.
Regulatory authority
1 · emerging
Standards and practice guides do not bind on their own, but GB 45438-2025 shows the committee's work carries mandatory force once adopted as a national standard.
analyst judgmentgb-45438-2025-aigc-labeling-standard
Standards influence
3 · central
Its generative AI security requirements and labeling practice guide pair directly with the binding CAC rules, and GB 45438-2025 names TC260 as the executing technical committee.
Technical capacity
2 · established
Its published documents specify corpus security, model security, and security-assessment requirements at working technical depth.
Convening power
2 · established
The WG9 formation drew a frontier lab into committee leadership, with SHLAB's director heading the group.
analyst judgmentshlab-wg9-formation-2026
International interfacenot assessed
What safety means here
Leadership
No leadership roles are recorded with a verifiable source.
History
- 2023-08-25
Practice guide on labeling methods for generative AI content published.
- 2024-02-29
Basic security requirements for generative AI services published.
- 2024-09-09
AI Safety Governance Framework released.
- 2025-02-28
GB 45438-2025, the mandatory labeling-method standard executed by TC260, issued.
- 2026-04-09
AI Security Standards Working Group (WG9) formed under a TC260 notice, with SHLAB's director as group head.
Connections
Recorded edges run downward and sideways: TC260 established WG9, and its secretariat operations show ties to CESI. analyst judgment
Affiliated With
- China Electronics Standardization Institute TC260 shows an operational affiliation with CESI through its secretary-general placement and homepage contact-location signals, though the retrieved sources stop short of a formal administrative-subordination statement.
Establishes
- AI Security Standards Working Group SHLAB's official article states that, according to a TC260 notice, AI Security Standards Working Group WG9 was formally established.
Selected outputs
Engagement fit
Assessmentanalyst judgment
- Plausible for
- Technically literate, standards-centered dialogue on AI security requirements and evaluation methods.
- Constrained by
- Standards influence is not enforcement power; binding decisions sit with the regulators its documents support.
Open questions
- A dedicated TC260 source explicitly naming the secretariat institution would firm up the administrative link to CESI.
- Whether WG9 output hardens quickly into national standards or stays at practice-guide level.
Source appendix (8 records)
- tc260-genai-service-security-requirements-2024
Cybersecurity Technology Basic Security Requirements for Generative AI Services · National Cybersecurity Standardization Technical Committee · 2024-02-29open
- tc260-content-labeling-practice-guide-2023
Cybersecurity Standards Practice Guide Method for Labeling Generated and Synthesized Content in Generative AI Services · National Cybersecurity Standardization Technical Committee · 2023-08-25open
- gb-45438-2025-aigc-labeling-standard
GB 45438-2025 Cybersecurity Technology: Labeling Method for AI-Generated and Synthetic Content · State Administration for Market Regulation / Standardization Administration of China · 2025-02-28open
- shlab-wg9-formation-2026
SHLAB Takes the Lead in Establishing the AI Security Standards Working Group and Holds Its First Meeting · Shanghai Artificial Intelligence Laboratory · 2026-04-09open
- tc260-ai-safety-governance-framework-2024
AI Safety Governance Framework · National Cybersecurity Standardization Technical Committee · 2024-09-09open
- report-china-standards-evaluation-2026-06-05
China Standards and Evaluation Research Brief · Asia AI Safety Atlas research brief
- tc260-homepage
National Cybersecurity Standardization Technical Committee Homepage · National Cybersecurity Standardization Technical Committee · unknownopen
- tc260-committee-roster-2024
Fourth Committee Roster of the National Cybersecurity Standardization Technical Committee · National Cybersecurity Standardization Technical Committee · 2024-04-03open
Last verified 2026-07-07