The following finalized Chinese national standard is designed to improve the safety and security of generative AI services. Relative to an earlier draft of the standard—also translated by CSET—this final version adds much more detail on best practices such as establishing procedures to ensure model training data do not include personally identifiable information or copyrighted works unless the model developer has explicit permission to do so. However, it is unclear how enforceable the more stringent measures of this standard are, given the cutthroat competition and rapid model iteration in the Chinese AI industry.

An archived version of the Chinese source text is available online at:    https://perma.cc/2JRP-4VDD.

National Standard of the People’s Republic of China

Cybersecurity Technology – Basic Safety Requirements for Generative Artificial Intelligence Services¹

GB/T 45654—2025

Introduction

At present, generative artificial intelligence (GenAI) technologies are undergoing continuous development, and related services have been widely applied, providing convenience across various aspects of social production and daily life. At the same time, however, they have also given rise to a large number of new cybersecurity risks and challenges, making it urgently necessary to establish safety baselines through standards and specifications.

This document primarily targets generative AI services that possess public opinion–shaping attributes or social mobilization capabilities, and it supports the conduct of work in areas such as filing and registration management, testing, and evaluation. When the focus is on data annotation safety, this document may be used in conjunction with Cybersecurity Technology—Generative Artificial Intelligence Data Annotation Safety Specifications (GB/T 45674); when the focus is on the safety of pre-training and fine-tuning data, this document may be used in conjunction with Cybersecurity Technology—Security Specifications for Generative Artificial Intelligence Pre-Training and Fine-Tuning Data (GB/T 45652).

1 Scope

This document² specifies requirements for GenAI services in areas including training data safety, model safety, and safety measures.³

This document applies to service providers conducting activities related to GenAI services and also serves as a reference for the relevant main oversight departments (主管部门) and third-party evaluation institutions.

Note: The major safety risks related to training data and generated content are provided in Appendix A, and reference methods for safety assessment of GenAI services are provided in Appendix B.

2 Normative Reference Documents

The contents of the following documents, through normative references in this text, constitute indispensable provisions of this document. Among them, for dated references, only the edition corresponding to that date applies to this document. For undated references, the latest edition (including all amendments) applies to this document.

GB/T 25069 Information Security Technology Terminology

To view the rest of this translation, download the pdf below.

1. Translator’s note: CSET translated an earlier (May 2024) draft of this standard. This translation is available online at: https://cset.georgetown.edu/publication/china-gen-ai-safety-standard-draft/.
2. Translator’s note: The authors of this standard formulated it based on technical documentation on AI safety published by SAC/TC260 in February 2024. An English translation of this technical documentation is available on CSET’s website at: https://cset.georgetown.edu/publication/china-safety-requirements-for-generative-ai-final/.
3. Translator’s note: The Chinese word 安全 ānquán—found in the title of this standard and throughout its text—can be translated into English as either “safety” or “security.” The Chinese authors of this standard provided the following English translation of its title: “Basic security requirements for generative artificial intelligence service.” However, this CSET English translation renders 安全 as “safety” in most cases, because in the context of this standard, the authors are mainly discussing the prevention of accidents or unforeseen problems (“safety”) of generative AI, rather than the prevention of deliberate abuse or sabotage (“security”).