Research on the Application of the Safe Harbor Rule to Generative Artificial Intelligence Service Providers -- The Construction of a Hierarchical Necessary Measures System
Abstract
to exempt themselves from tort liability for generated content. Legally, GAI service providers qualify as network service providers, allowing
direct application of the rule. From a policy perspective, this application stimulates technological innovation and protects user rights. Referencing the EU Artificial Intelligence Act's risk classification, GAI systems are divided into four levels: "minimal risk, " "limited risk, " "high
risk, " and "unacceptable risk." Corresponding exemption conditions apply: minimal-risk systems fulfill post-event response obligations;
limited-risk systems add risk notification and generation identification; high-risk systems assume process supervision and model optimization;
unacceptable-risk systems are ineligible for exemption.
Keywords
Full Text:
PDFReferences
[1] Liu Xiao, Ye Yuhao. The Application Dilemma and Solution Path of China's "Safe Harbor Principle" under the Background of the Civil
Code. Electronic Intellectual Property. 2022; (5): 28-38.
[2] Zhang Lili, Hua Zhiyu. Legal Risks and Regulation of Generative Artificial Intelligence. Administration and Law. 2024; (4): 88-100.
[3] Xu Wei. On the Legal Status and Liability of Generative Artificial Intelligence Service Providers. Science of Law (Journal of Northwest
University of Political Science and Law). 2023; 41(4): 69-80.
[4] Zhou Xuefeng. An Analysis of the Tort Liability of Generative Artificial Intelligence. Journal of Comparative Law. 2023; (4): 117-131.
[5] Chen Zhenni. The Tort Liability of Generative Artificial IntelligenceReflection from the Perspective of Human-Computer Interaction. Western Law Review. 2024; (4): 23-34.
[6] Yuan Xiuting, Wang Han. Copyright Regulation of Infringing Content Generated by Artificial Intelligence. Electronic Intellectual Property. 2025; (3): 4-16.
[7] Browne R. SAP Boss Warns Against Regulating AI, Says Europe Risks Falling Behind U.S., China. Available from: https://www.cnbc.
com/2024/10/22/sap-ceo-urges-europe-not-to-regulate-ai-says-will-put-region-behind.html [Accessed 22nd October 2024].
[8] Zhao Zhonghua, Zhang Qian. Getting Out of the "Collingridge Dilemma": Risk Representation and Governance Exploration of Generative Artificial Intelligence. Available from: http://kns.cnki.net/kcms/detail/21.1383.C.20250616.0925.006.html [Accessed 8th July 2025].
[9] European Parliament and the Council of the European Union. REGULATION (EU) 2024/1689 (Artificial Intelligence Act). 2024.
[10] Youku Network Technology (Beijing) Co., Ltd. v. Beijing Baidu Netcom Technology Co., Ltd. Beijing Intellectual Property Court, (2020)
Jing 73 Min Zhong No. 155 Civil Judgment. 2020.
DOI: http://dx.doi.org/10.70711/aitr.v3i6.8591
Refbacks
- There are currently no refbacks.
