The following Chinese draft national standard proposes safety and security rules for the training and fine-tuning data used to develop generative AI models. The standard defines safety and security as including not only protection of people’s physical safety and disinformation prevention, but also censorship of content that criticizes Communist Party rule or presents China in an unflattering light. China issued a finalized version of these standards in April 2025, but, as of the publication date of this translation, CSET has not observed a publicly available full-text copy of the final version.

An archived version of the Chinese source text is available online at:    https://perma.cc/YYU8-6DH4

National Standard of the People’s Republic of China

Cybersecurity Technology – Safety Specifications for Generative Artificial Intelligence Pre-Training and Fine-Tuning Data

(Draft for Feedback)

(Draft Completed on: March 28, 2024)

1. Scope

This document specifies the safety¹ requirements for generative artificial intelligence (GenAI) pre-training and fine-tuning data and related processing activities, and describes the corresponding evaluation methods.

This document applies to guiding GenAI service providers in carrying out pre-training and fine-tuning data processing activities, as well as in conducting self-evaluations of the safety of pre-training and fine-tuning data. It may also serve as a reference for regulatory assessments.

2. Normative References

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 AAAAA Cybersecurity Technology – Generative Artificial Intelligence Data Annotation Safety Specifications²

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

1. Translator’s note: The Chinese word 安全 encompasses the meanings of both “safety” (protection from accidental harm) and “security” (protection from deliberate harm). In this translation, it is variously translated as “safety,” “security,” “safety and security,” or “safety or security” at the translator’s discretion.
2. Translator’s note: CSET’s English translation of the draft version of the Chinese national standard Cybersecurity Technology – Generative Artificial Intelligence Data Annotation Safety Specifications is available online at: https://cset.georgetown.edu/publication/china-gen-ai-data-labeling-safety-standard-draft/. China issued a finalized version of this standard in April 2025, but, as of the publication date of this translation, CSET has not observed a publicly available full-text copy of the final version.