Skip to main content
Framework Authoritative — Peer-reviewed venues, standards bodies, regulators, official government publications

网络安全标准实践指南——人工智能应用安全指引 教育 (TC260-PG-20269A) [Practice Guide for Cybersecurity Standards: Security Guidelines for Artificial Intelligence Applications: Education]

Sector practice guide (23 pages) released together with the general AI Application Security Guidelines and companion guides for health and for broadcasting and online audiovisual services. It sets general and scenario-specific security requirements for AI use by educational institutions across teaching assistance, learning assistance, research assistance, administration and assessment. It contains a minors-protection mechanism, age-banded rules for student use of generative tools, supervision tooling for teachers and guardians, and requirements for handling high-risk content such as self-harm and psychological manipulation.

Publisher

National Technical Committee 260 on Cybersecurity of SAC (TC260) Secretariat; drafted with the Ministry of Education Department of Science, Technology and Informatization, the MoE Education Management Information Center, CESI, Beijing Normal University, Tsinghua University and East China Normal University among others

Published

15 Sept 2026

Added

today

DOI

Key Findings

  • Clause 5(e): AI services for minors must use age-appropriate notices and default protective settings, obtain guardian consent, keep minors' AI-assisted learning under guardian or teacher supervision, and provide usage-duration limits, function limits, guardian settings and anomaly alerts
  • Clause 6.2: primary-school pupils should not use open-ended generation alone; platforms must offer duration, frequency and time-of-day management (continuous-use limits, rest intervals, night-time limits) and restrict personalised recommendation, anthropomorphic interaction and paid features, with usage statistics and alerts for teachers or guardians
  • Clause 6.4(c): when input or output involves personal injury, behavioural inducement, psychological manipulation, cyberbullying or other risks to students' physical or mental health, the application must restrict answers, pause the interaction or give safety guidance according to risk level, alert teachers, guardians or the institution, and escalate life-safety risks to competent units
  • Annex A.2 classifies learning platforms with anthropomorphic characters or persistent emotional interaction as an anthropomorphic-interaction scenario and lists overuse, emotional dependence and excessive data collection among its risks
  • Annex A.4(d) limits AI in mental-health education to general knowledge, general emotional support and risk-cue identification with referral to qualified staff; it states AI does not replace professionals for diagnosis, treatment or crisis handling, and names misjudgment, delayed professional intervention, leakage of sensitive psychological data and student emotional dependence as the main risks
  • Clause 6.1: AI answers may not be given to students as final answers, teachers set the timing and frequency of AI involvement, and students must be able to end a session themselves

Methodology Notes

Soft-law practice guide from the TC260 secretariat (the document states it is a technical document accompanying standards, not a national standard). Chinese-language PDF, 23 pages, masthead TC260-PG-20269A v1.0-202609, PDF creation date 15 September 2026; released on tc260.org.cn with three sibling guides (general principles TC260-PG-20268A, health TC260-PG-202610A, broadcasting and online audiovisual TC260-PG-202611A). Read in full from the PDF; the clause summaries above are the sweep's translation, not an official English text. The PDF URL contains Chinese characters and must be percent-encoded for automated fetches.

Tags

tc260chinaeducationpractice-guideminorsanthropomorphic-interactionsoft-law

Cite This

APA

National Technical Committee 260 on Cybersecurity of SAC (TC260) Secretariat; drafted with the Ministry of Education Department of Science, Technology and Informatization, the MoE Education Management Information Center, CESI, Beijing Normal University, Tsinghua University and East China Normal University among others. (2026). 网络安全标准实践指南——人工智能应用安全指引 教育 (TC260-PG-20269A) [Practice Guide for Cybersecurity Standards: Security Guidelines for Artificial Intelligence Applications: Education]. https://www.tc260.org.cn/tc260/tzgg/202609/ec149159dcb246268eda751eee89967d/files/%E3%80%8A%E7%BD%91%E7%BB%9C%E5%AE%89%E5%85%A8%E6%A0%87%E5%87%86%E5%AE%9E%E8%B7%B5%E6%8C%87%E5%8D%97%E2%80%94%E2%80%94%E4%BA%BA%E5%B7%A5%E6%99%BA%E8%83%BD%E5%BA%94%E7%94%A8%E5%AE%89%E5%85%A8%E6%8C%87%E5%BC%95%20%E6%95%99%E8%82%B2%E3%80%8B.pdf.pdf