中国, 作为人工智能应用的前沿国家, 对生成式人工智能这一突破性技术的安全治理尤为重视。随着AI在各行业的广泛应用, 中国政府不仅高度重视AI的发展与安全, 而且自2017年起陆续出台了一系列相关规制。从≪中华人民共和国网络安全法≫到2023年实施的≪生成式人工智能服务管理暂行办法≫, 特别是2020至2023年这三年集中出台的法规, 旨在规范AI应用, 保障国家安全、社会公共利益和公民个人权益, 同时推动行业的健康发展。这些法律法规不仅反映了中国政府对AI安全治理的重视, 也展示了中国在AI领域法治化建设的决心和成就。 本研究通过对相关法律法规和对司法与政府相关部门监管的实践案例的梳理与分析, 旨在研究中国对生成式人工智能安全治理的规制特点和治理方式, 探讨其法律框架、政策导向及实践成效。中国在生成式人工智能安全治理上采取了以促进发展为导向, 形成政府与产业协同治理的模式。选择协同治理的原因, 一方面是科技快速发展和其复杂性, 以及法律的滞后性；另一方面, 单靠政府的法律法规治理可能会与产业形成零和博弈的局面, 这既不利于安全治理, 也不利于产业的创新发展。 本研究还关注了这些法律法规在应对生成式人工智能安全风险、促进AI健康发展方面的作用与不足, 以期为未来的法律法规制定和进一步完善提供有益的参考。 通过文献综述的研究方法, 本研究总结了中国对生成式人工智能治理已形成的初步框架和体系。尽管技术快速发展和法律的滞后性使得现有法律规制难以全面覆盖, 但中国的事前预防、事中管理和事后处理的治理方式具有实践参考价值。因此, 未来中国对生成式人工智能的规制政策需要结合技术的快速发展、市场应用以及吸引外资等现实需求, 不断对现有法律规制进行完善。 本研究通过对中国生成式人工智能安全治理法律法规的研究, 深入了解了中国在这一领域的法治化建设进程, 期望为其他国家和地区在人工智能安全治理方面提供借鉴和启示, 推动全球生成式人工智能健康、有序、可持续的发展, 发挥该研究的理论和实践参考价值。

China, at the forefront of artificial intelligence (AI) application, prioritizes the safety governance of generative AI, a revolutionary technology. As AI permeates various sectors, the Chinese government has placed significant emphasis on both the development and security of AI, enacting a suite of regulations since 2017. This includes the “Cybersecurity Law of the People’s Republic of China” and the “Interim Measures for the Management of Generative AI Services” implemented in 2023. Particularly noteworthy is the flurry of legal activity from 2020 to 2023, aimed at standardizing AI applications to safeguard national security, public welfare, and individual rights, while fostering the industry’s healthy growth. These legislative efforts not only underscore the Chinese government’s dedication to AI safety governance but also highlight China’s resolve and progress in establishing a legal framework for AI. This research endeavors to delve into the distinctive regulatory features and governance approaches of China’s management of generative artificial intelligence safety. It conducts a comprehensive analysis of pertinent legal and regulatory frameworks, along with case studies of oversight by judicial and governmental bodies, to examine the efficacy of its policies and practices. China has adopted a development-oriented strategy for generative AI safety governance, fostering a collaborative governance model between government and industry. This collaborative approach is chosen to cope with the rapid technological advancements and their inherent complexities, as well as the inherent lag of legal systems. Moreover, relying solely on government regulation could result in a zero-sum game with the industry, which is detrimental to both safety governance and the industry’s innovative development. The paper also examines the effectiveness and deficiencies of these laws and regulations in managing the safety risks of generative AI and in promoting its healthy development, aiming to offer insights for future legislative development and refinement. Employing a literature review methodology, this paper synthesizes that China has established a preliminary framework and system for the governance of generative AI. Although the rapid pace of technological advancement and the lag of legal systems mean that current regulations cannot cover every aspect, China’s governance approach, which includes preemptive measures, ongoing management, and post-incident handling, holds valuable practical insights. Therefore, future Chinese policies on generative AI should be continuously refined in light of technological advancements, market applications, and the need to attract foreign investment. By delving into the legal regulations governing China’s safety governance of generative AI, this paper provides an in-depth look at China’s progress in legalizing this domain, offering a reference and inspiration for other nations and regions in their AI safety governance endeavors. This is intended to foster the healthy, orderly, and sustainable development of global generative AI, thereby enhancing the theoretical and practical value of this research.