China GB/T 45654-2025 Basic Security Requirements for Generative AI Services
China's national standard for generative AI service security: the training data requirements (source sampling with a five per cent threshold, diversity, provenance and licences, corpus filtering, intellectual property and personal information protection, annotator qualification, annotation rules, accuracy review and isolated storage), the model requirements (content security as a training objective, framework audits and backdoor detection, a ninety per cent output pass rate, refusal behaviour, synthetic content labelling, input monitoring, reassessment after major updates, training and inference isolation) and the security measures (use-case justification, high-risk scenario and minor protection, transparency, opt-out from training on user input, complaints, abusive input handling and monitoring staff, continuity and on-device model security), with the 31-risk taxonomy and the assessment method.
China GB/T 45654-2025 Basic Security Requirements for Generative AI Services is a compliance framework from China with 3 domains and 22 controls. The largest domains are Clause 4: training data security (sources, content, annotation) – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services (10 controls), Clause 6: security measures – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services (7 controls), Clause 5: model security – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services (5 controls). Every control below carries what it requires and what an assessor expects to see.
Framework summaries on this platform are interpretations grounded in the published source, for educational and compliance planning purposes. They do not reproduce or replace the official standards. Refer to the authoritative source for the definitive text. Framework names and trademarks belong to their respective organisations.
Framework Domains (3)
Clause 4: training data security (sources, content, annotation) – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services
| Code | Title |
|---|---|
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.1.1 | Training data source selection: sample before and after collection, reject sources above five per cent illegal or unhealthy content |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.1.2 | Diversity and mix of training data sources |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.1.3 | Management and traceability of training data sources: licences, collection records, commercial contracts and user authorisation |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.2.1 | Filter all training data of every type before training |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.2.2 | Intellectual property protection in training data and generated content |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.2.3 | Personal information in training data only with consent or another lawful basis |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.3.1 | Annotation personnel: training, assessment, qualification and separation of duties |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.3.2 | Annotation rules for functional and security annotation |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.3.3 | Accuracy of annotated content: batch sampling, re-annotation and review |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::4.3.4 | Isolated storage of security annotation data |
Clause 5: model security – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services
| Code | Title |
|---|---|
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::5.1 | Model training security: content security as a core optimisation metric, framework audits and backdoor detection |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::5.2 | Model output security: ninety per cent pass rate, accuracy, reliability, refusal behaviour and content labelling |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::5.3 | Model monitoring and evaluation: input monitoring, routine testing and emergency handling |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::5.4 | Model update and upgrade security: policy and self-assessment after major changes |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::5.5 | Isolation of training and inference environments |
Clause 6: security measures – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services
| Code | Title |
|---|---|
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::6.1 | Service suitability: justify use cases, protect high-risk scenarios and protect minors |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::6.2 | Service transparency: publish audience, scenarios, purposes, limitations, model summary and personal information use |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::6.3 | User input used for training only with a prominent notice and an easy opt-out |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::6.4 | Public and user complaint and reporting channels with rules and time limits |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::6.5 | Serving users: input detection with published suspension rules, and adequate monitoring staff |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::6.6 | Service stability and continuity: backup and recovery |
| china-gb-t-45654-2025-basic-security-requirements-for-generative-ai-services::6.7 | On-device models: activation, on-device security module, keyword updates and model patching |
What is China GB/T 45654-2025 Basic Security Requirements for Generative AI Services and who does it apply to?
China GB/T 45654-2025 Basic Security Requirements for Generative AI Services is a compliance framework from China with 3 domains and 22 controls. China's national standard for generative AI service security: the training data requirements (source sampling with a five per cent threshold, diversity, provenance and licences, corpus filtering, intellectual property and personal information protection, annotator qualification, annotation rules, accuracy review and isolated storage), the model requirements (content security as a training objective, framework audits and backdoor detection, a ninety per cent output pass rate, refusal behaviour, synthetic content labelling, input monitoring, reassessment after major updates, training and inference isolation) and the security measures (use-case justification, high-risk scenario and minor protection, transparency, opt-out from training on user input, complaints, abusive input handling and monitoring staff, continuity and on-device model security), with the 31-risk taxonomy and the assessment method. It is used by organisations to establish and maintain compliance with industry standards and regulatory requirements.
What does China GB/T 45654-2025 Basic Security Requirements for Generative AI Services actually require?
China GB/T 45654-2025 Basic Security Requirements for Generative AI Services has 22 controls organised across 3 domains. The largest domains are Clause 4: training data security (sources, content, annotation) – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services (10 controls), Clause 6: security measures – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services (7 controls), Clause 5: model security – China GB/T 45654-2025 Basic Security Requirements for Generative AI Services (5 controls). Each control defines specific requirements that organisations must implement to achieve compliance.
If I already comply with another framework, how much of China GB/T 45654-2025 Basic Security Requirements for Generative AI Services do I already cover?
China GB/T 45654-2025 Basic Security Requirements for Generative AI Services does not currently have cross-framework mappings in our system. Check back as we continuously expand our mapping database.
How do I implement China GB/T 45654-2025 Basic Security Requirements for Generative AI Services?
Start your China GB/T 45654-2025 Basic Security Requirements for Generative AI Services compliance journey by running a self-assessment on our platform to identify your current compliance posture. Our AI advisory can answer specific questions about China GB/T 45654-2025 Basic Security Requirements for Generative AI Services requirements, and cross-framework mapping helps you leverage existing controls from other frameworks you may already comply with. Create a free account to access all 22 controls and track your progress.
Start Your Compliance Journey
Create a free account to run self-assessments, get AI advisory, and track your compliance progress across 1,248 frameworks.
Get Started Free →Free forever, no credit card required