A Control-Based Differentiated Framework for the Content Governance Obligations of Generative Artificial Intelligence Service Providers

Shengtao Tai

doi:10.62051/ijsspa.v10n8.03

Authors

Shengtao Tai Law School, Beijing Normal University, Beijing, China

DOI:

https://doi.org/10.62051/ijsspa.v10n8.03

Keywords:

Generative Artificial Intelligence, Content Governance Obligations, Duty of Care, Risk-control Theory

Abstract

The content governance obligations of generative artificial intelligence (GAI) service providers constitute a central issue in contemporary AI regulation. Existing legal frameworks, however, impose substantially uniform content governance obligations on GAI service providers without adequately accounting for the significant differences among various service models. Drawing upon risk-control theory, this Article argues that the content governance obligations of GAI service providers should be differentiated according to the degree of substantive control they exercise over the content-generation process. Based on differences in such control, GAI service models may be categorized into three types: pure prompt-based generation services, retrieval-augmented generation (RAG) services, and foundation model fine-tuning services. Correspondingly, each category should be subject to a distinct set of content governance obligations calibrated to its respective level and form of control. A differentiated regulatory framework that allocates content governance obligations on the basis of substantive control can better reconcile the dual objectives of safeguarding the security of AI-generated content and promoting technological innovation, thereby achieving a more proportionate and refined balance between regulation and innovation.

Downloads

Download data is not yet available.

References

[1] Wu, H. D. (2025). On the reasonable duty of care of generative AI service providers. Intellectual Property, (11), 3–23. [In Chinese.]

[2] Wu, Y. R. (2025). The liability allocation mechanism for generative AI service providers. Academic Research, (10), 60–68. [In Chinese.]

[3] Liu, X. L. (2026). Legal clarification and path optimization of the obligation to label AI generated content. Beijing Social Sciences, (1), 105–115. [In Chinese.]

[4] Cheng, X. (2026). Duty of care, statutory duty, and fault of generative AI service providers. Political Science and Law Tribune, (4), 35–47. [In Chinese.]

[5] Zhang, W. R., & Liu, Y. Z. (2026). Hierarchical construction of imputation principles for generative AI service providers. Hebei Law Science, (7), 178–194. [In Chinese.]

[6] Guo, M. R. (2016). The impact of dangerous liability on the development of tort liability law. Journal of Yantai University (Philosophy and Social Sciences Edition), (1), 24–29. [In Chinese.]

[7] Liu, C. W., & Ma, R. C. (2025). Normative sources of the duty of care for generative AI users. Journal of Shandong University (Philosophy and Social Sciences Edition), (5), 152–163. [In Chinese.]

[8] Zhang, X. B., & Bian, L. (2025). On the presumption of fault liability of generative AI service providers. Northern Legal Science, (5), 5–19. [In Chinese.]

[9] Liu, Y. H. (2026). Research on hierarchical allocation of copyright infringement liability for generative AI. Law Journal, (3), 67–87. [In Chinese.]

[10] He, W. (2024). Generative AI AIGC and multimodal technology application practice guide. China Science and Technology Press. [In Chinese.]

[11] He, Y. P. (2026). On the liability of model providers for copyright infringement of AI generated content. Journal of Soochow University (Law Edition), (1), 51–63. [In Chinese.]

[12] Li, Z. H. M. (2026). Secondary hallucination risks and regulation of retrieval augmented generation technology in judicial applications. Journal of University of Electronic Science and Technology of China (Social Sciences Edition). Advance online publication. https://doi.org/10.14071/j.1008 8105(2026) 3013. [In Chinese.]

[13] Tian, Y. L., Wang, Y. T., Wang, X. X., Yang, J., Shen, T. Y., Wang, J. G., Fan, L. L., Guo, C., Wang, S. W., Zhao, Y. W., Wan, S., & Wang, F. Y. (2025). From RAG to SAGE: Current status and prospects. Acta Automatica Sinica, (6), 1145–1169. [In Chinese.]

[14] Wang, X. L., Li, Y., Ma, C. F., & Li, S. (2026). Retrieval augmented generation (RAG): A survey of methods and applications. Computer Science, (7), 101–117. [In Chinese.]

[15] Tang, A. D., & Lin, Z. C. (2026). A survey on parameter efficient fine tuning methods for large models: Techniques, trends, and challenges. Acta Automatica Sinica. Advance online publication. https://doi.org/10.16383/j.aas.c250451. [In Chinese.]

[16] Qin, D. H., Li, Z. T., Bai, F. B., Dong, L. K., Zhang, H., & Xu, C. (2025). A survey on parameter efficient fine tuning techniques for large language models. Computer Engineering and Applications, (16), 38–63. [In Chinese.]

[17] Hu, H. F., & Zhou, Y. (2024). Multi subject personalized image generation method based on LoRA. Chinese Journal of Stereology and Image Analysis, (4), 320–332. [In Chinese.]

[18] Ning, D. X., & Wei, T. T. (2026). Hierarchical determination of indirect infringement and liability allocation of generative AI service providers: Also on the imputation dilemma and response of indirect infringement of traditional network service providers. Social Sciences in Inner Mongolia, (1), 96–103. [In Chinese.]

[19] Du, M. J. (2025). Systematic deconstruction and institutional reconstruction of tort liability of generative AI service providers. Journal of Xihua University (Philosophy and Social Sciences Edition), (4), 25–39. [In Chinese.]

[20] Wang, L. M., & Lin, T. L. (2026). Judicial regulation of generative AI: Risks, dilemmas, and responses. Fujian Tribune (Humanities and Social Sciences Edition), (4), 143–154. [In Chinese.]

[21] Zhang, H. T. (2025). Imputation and composition of generative AI torts under the dynamic system theory. Law Journal, (6), 82–101. [In Chinese.]

[22] Li, D. (2023). On the determination of tort liability for algorithmic discrimination against consumers: An empirical investigation based on judicial adjudication. Contemporary Law Review, (6), 75–85. [In Chinese.]

[23] Zhang, L. H. (2023). The legal positioning and hierarchical governance of generative AI. Modern Law Science, (4), 126–141. [In Chinese.]

[24] Deng, W. (2026). Research on the regulatory path of generative AI hallucination infringement. Political Science and Law, (6), 134–150. [In Chinese.]

[25] Li, D. (2026). Phased duty of care and tort liability of generative AI service providers. Modern Law Science, (2), 224–242. [In Chinese.]

[26] Wang, Q. S. (2025). Attribution paradigm and liability mechanism for damages caused by large model operations. Oriental Law, (4), 124–137. [In Chinese.]