Every AI Policy Ever · open in the interactive register

Published (non-binding, current)Non-binding · Guidelines / framework

Cybersecurity Standard Practice Guide — Guide to Security Emergency Response for Generative Artificial Intelligence Services

网络安全标准实践指南——生成式人工智能服务安全应急响应指南

Record CN-25 · TC260-PG-202515A

Issued by
National Cybersecurity Standardization Technical Committee (TC260) Secretariat
Jurisdiction
China
Level · region
National · East Asia
Published / adopted
2025-09-15
Language
Chinese
Official identifier
TC260-PG-202515A

Summary

TC260 practice guide that implements the Interim Measures for generative AI services by giving providers a method to classify and grade generative AI service security incidents and an emergency response process. Incidents are information content, data security, cyberattack or other incidents, graded from level 1 (especially major) to level 4 (general). The process covers preparation (response team, plans, keyword and test-question banks, drills), monitoring (a refusal rate of at least 95% on the should-refuse test bank for hot events), handling, with service restriction or takedown and reporting of serious incidents to the competent department, and review.

Language

No English version of this instrument was found. The English title and the summary are the register's own, written from the sources below; for the exact wording use the Chinese text.

Safety topics

incident reportinggeneral-purpose / frontier AIevaluation & testing

Status notes

Version V1.0-202509, released with notice 网安秘字〔2025〕125号 dated 15 Sept 2025 and posted on 22 Sept 2025. A generative AI service security incident is defined as an event caused by human factors, malicious attacks or inherent model flaws that damages the service or its data and has negative effects on the state, society or economy. Classification and grading follow GB/T 20986-2023, and illegal and unhealthy information mainly means the 29 risks in Annex A (A.1 to A.4) of GB/T 45654-2025 (CN-10). Self-assessed incidents of level 3 or higher with possible wide public impact are to be reported to the competent department immediately when they occur five times within 24 hours, affect more than 10,000 users and may cause adverse social impact; general incidents are handled internally and reported periodically. Annex A gives cases of a discrimination incident and a false-information incident.

Sources

Open the official source https://www.tc260.org.cn/tc260/sjzn/202509/dbafd382ac164506b84065f37f2cebca/files/0d784893a2fe42f396e88d95317a787c.pdf

Secondary source: https://www.tc260.org.cn/tc260/tzgg/202509/ace43eefdec6497cb6ee7510dc3aa07a.shtml

Local copy

Archived. Last attempt 2026-10-18 00:45:29 UTC

Official source (third-party copy, not published) archive/_runs/2026-10-14-discover-leads-2026-10-14/C009/2026-10-14/official.pdf
Retrieved 2026-10-14 13:10:52 UTC
SHA-256 35d572a8d8072d21ca6f9cefdfd9316d1fad96f4c1b8f24ca1a53239c71b41cc
Secondary source (third-party copy, not published) archive/CN-25/2026-10-18/secondary.html
Retrieved 2026-10-18 00:13:36 UTC
SHA-256 8da8d09c96fe37e952159516e655ca0c8944ce964db6c88e6c01ba761057dcac

Verification

B
Another official source, such as an announcement by the issuer, confirms the title, date and status. checked 2026-10-18 · alternative-official-url-fetched
Recheck of 18 Oct 2026. Read the guide's PDF from tc260.org.cn (version V1.0-202509; stored 14 Oct 2026, and identical when fetched again on 18 Oct 2026) and TC260's release notice 网安秘字〔2025〕125号 (tc260.org.cn, stored 18 Oct 2026 as the record's secondary source). The PDF confirms the title, the number TC260-PG-202515A, the TC260 Secretariat as issuer, the version and the contents; its cover gives only September 2025 (2025年9月). The notice, signed on 15 Sept 2025 and posted on 22 Sept 2025, gives the day and says the guide implements the Interim Measures for generative AI services. The summary follows the Chinese text. No English version was found; the guide's statement reserves translation to the TC260 Secretariat's written authorisation.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-18: 7 field(s) supported.

History

  1. 2026-10-18verification.level: A → B — level computed from the evidence by tools/pipeline.py2026-10-18-recheck-east-asia-1 · Claude Code (claude-opus-5-5)
  2. 2026-10-14related: [{"id": "CN-03", "slug": "cn-generative-ai-interim-measures"}, {"id": "CN-10", " → [{"id": "CN-03", "slug": "cn-generative-ai-interim-measures"}, {"id": "CN-10", " — TC260 practice guides of one series or referring to each other, as the checker of run 2026-10-14-discover-leads-2026-10-14 named them (checks/C003.json; C011's own text refers to TC260-PG-20266A)2026-10-14-discover-leads-2026-10-14 · Claude Code (claude-opus-5-5)
  3. 2026-10-14Added to the register — found in pipeline run 2026-10-14-discover-leads-2026-10-14, candidate C009 (CN/ministries)2026-10-14-discover-leads-2026-10-14 · Claude Code (claude-opus-5-5)

Related records

Generative AI Measures China · In force

GB/T 45654-2025 China · Published (non-binding, current)

TC260-PG-20268A (AI application security, general provisions) China · Published (non-binding, current)

Citation

National Cybersecurity Standardization Technical Committee (TC260) Secretariat. Cybersecurity Standard Practice Guide — Guide to Security Emergency Response for Generative Artificial Intelligence Services [TC260-PG-202515A]. 2025-09-15. https://www.tc260.org.cn/tc260/sjzn/202509/dbafd382ac164506b84065f37f2cebca/files/0d784893a2fe42f396e88d95317a787c.pdf. Status: Published (non-binding, current); verified 2026-10-18 (level B). Local copy archive/_runs/2026-10-14-discover-leads-2026-10-14/C009/2026-10-14/official.pdf retrieved 2026-10-14 13:10:52 UTC, SHA-256 35d572a8d8072d21ca6f9cefdfd9316d1fad96f4c1b8f24ca1a53239c71b41cc. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record CN-25.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.