|
|
|
| Construction of Regulatory Sandbox for Generative AI |
| Luo Zhimin, Hong Yifan |
| Law School, China University of Political Science and Law, Beijing 100088, China |
|
|
|
|
Abstract The rapid development of generative AI has brought both technological revolution and multiple security risks. How to balance risk regulation and innovation incentives has become an important issue. Regulatory sandbox, as an innovative regulatory tool, responds to the requirements of inclusive and prudential regulation and the concept of agile governance. However, domestic research on regulatory sandboxes mainly focuses on the field of Fintech, and there is no research on regulatory sandbox for AI, especially for generative AI. This study aims at exploring the effective regulation and innovation incentives for generative AI through the construction of the regulatory sandbox for generative AI.Existing regulatory models, whether “command-and-control” or “development first, governance later”, have difficulties. The introduction of inclusive and prudential regulation and the concept of agile governance has become inevitable. Based on it, regulatory sandboxes for generative AI have feasibility in China. The regulatory sandbox for AI has been applied in more and more extra-territorial countries. Some countries are also exploring the construction of the regulatory sandbox for generative AI. The domestic regulatory sandbox has also accumulated a certain amount of experience, and the generative AI industry is also accelerating. It should be noted that AI involves innovative projects in different scenarios and fields, which means that a single, undifferentiated regulatory sandbox for AI may be more inappropriate. What’s more, the expression “regulatory sandbox” is more appropriate for regulatory sandbox for generative AI. In terms of governance effects, the construction of a generative AI regulatory sandbox is beneficial to multiple actors, reducing risks for innovators, improving regulation, and protecting consumers.The path to implementation of regulatory sandbox for generative AI is discussed next. To ensure that the regulatory sandbox has a legal basis, the plan to flexibly implement the contents set forth in laws, administrative regulations and departmental rules through experimental legislation within the reform pilot zone is permissible. Instead, the following two options should be adopted to make the regulatory sandbox legally enforceable. One is to make a decision by the National People’s Congress and its Standing Committee, and the second is to formulate a unified Artificial Intelligence law and write it into the regulatory sandbox. At the micro level, the entire regulatory process should be designed. First, at the application stage, the entry threshold should be determined and three categories and six principles should be constructed, namely, “substantial innovation”, “tilting towards small and micro enterprises”, and “risk regulation needs”, “project maturity”, “public interest principle” and “domestic market application principle”. Secondly, in the testing phase, clear exemption rules and effective communication are required to build exemption rules based on time, application scenarios and user types, as well as a mechanism for disclosure of information during the whole process and effective communication between regulators and innovators. Third, in the exit phase, planning the exit conditions, regulators need to make two types of five judgement results for different situations; after exit, regulators should require innovators to complete their exit obligations, properly handle data and personal information, and jointly prepare regulatory sandbox exit reports. It is important that we continue to explore and improve the regulatory sandbox for generative AI in practice so that generative AI can better promote China’s economic and social development.
|
|
Received: 17 December 2024
|
|
|
|
1 沈芳君:《生成式人工智能的风险与治理——兼论如何打破“科林格里奇困境”》,《浙江大学学报(人文社会科学版)》2024年第6期,第73-91页。 2 张涛:《生成式人工智能训练数据集的法律风险与包容审慎规制》,《比较法研究》2024年第4期,第86-103页。 3 康骁:《行政法如何应对生成式人工智能——基于算法、训练数据和内容的考察》,《云南社会科学》2024年第4期,第80-88页。 4 姚佳:《人工智能的训练数据制度——以“智能涌现”为观察视角》,《贵州社会科学》2024年第2期,第51-57页。 5 姚志伟、李卓霖:《生成式人工智能内容风险的法律规制》,《西安交通大学学报(社会科学版)》2023年第5期,第147-160页。 6 Financial Conduct Authority, “Regulatory sandbox,” 2015-11, https://www.fca.org.uk/publication/research/regulatory-sandbox.pdf, 2024-12-12. 7 胡滨、郑联盛等:《监管沙盒:理论框架与国际经验》,北京:中国金融出版社,2020年。 8 刘盛:《金融监管沙盒制度研究》,北京:法律出版社,2023年。 9 钟浩南:《人工智能伦理治理的元规制进路:以企业监管为焦点》,《江淮论坛》2024年第5期,第148-155页。 10 徐磊:《发展与安全并重:生成式人工智能风险的包容审慎监管》,《理论与改革》2024年第4期,第67-83,176页。 11 戴杕:《ChatGPT:人工智能的通用性发展及其法律规制》,《西南政法大学学报》2023年第4期,第86-97页。 12 丁晓东:《全球比较下的我国人工智能立法》,《比较法研究》2024年第4期,第51-66页。 13 戴昕:《人工智能治理的两种安全思维:批判与超越》,《环球法律评论》2025年第6期,第22-35页。 14 曹建峰:《迈向可信AI:ChatGPT类生成式人工智能的治理挑战及应对》,《上海政法学院学报(法治论丛)》2023年第4期,第28-42页。 15 佚名:《谷歌再吃罚单 法国指其聊天机器人侵犯版权》,2024年3月21日,https://news.cctv.com/2024/03/21/ARTIrfUj1GDIaYE2Bi5ovcOQ240321.shtml,2024年12月12日。 16 European Commission, “Commission compels Microsoft to provide information under the Digital Services Act on generative AI risks on Bing,” 2024-05-17, https://digital-strategy.ec.europa.eu/en/news/commission-compels-microsoft-provide-information-under-digital-services-act-generative-ai-risks, 2024-12-12. 17 张乐、赵爽爽:《人工智能的科林格里奇困境与安全监管的敏捷性调适》,《中国行政管理》2026年第3期,第38-46页。 18 Weatherbed J., “European companies claim the EU’s AI Act could ‘jeopardise technological sovereignty’,” 2023-06-30, https://www.theverge.com/2023/6/30/23779611/eu-ai-act-open-letter-artificial-intelligence-regulation-renault-siemens, 2024-12-12. 19 Coulter M. & Mukherjee S., “Behind EU lawmakers’ challenge to rein in ChatGPT and generative AI,” 2023-04-28, https://financialpost.com/pmn/business-pmn/behind-eu-lawmakers-challenge-to-rein-in-chatgpt-and-generative-ai, 2024-12-12. 20 胡凌:《“非法兴起”:理解中国互联网演进的一个视角》,《文化纵横》2016年第5期,第120-125页。 21 张志安、冉桢:《发展实用主义:中国互联网政策变迁与平台治理的内在逻辑》,《新闻与写作》2023年第1期,第89-97页。 22 程乐:《生成式人工智能的法律规制——以ChatGPT为视角》,《政法论丛》2023年第4期,第69-80页。 23 Helman L., “Curated innovation,” Akron Law Review, Vol. 49, No. 3 (2015), pp. 695-738. 24 任嵘嵘、齐佳丽、苏露阳:《敏捷治理:一个新的管理变革——研究述评与展望》,《技术经济》2021年第8期,第133-144页。 25 World Economic Forum, “Agile governance: reimagining policy-making in the fourth industrial revolution,” 2018-04-24, https://www.weforum.org/publications/agile-governance-reimagining-policy-making-in-the-fourth-industrial-revolution/, 2024-12-12. 26 薛澜、赵静:《走向敏捷治理:新兴产业发展与监管模式探究》,《中国行政管理》2019年第8期,第28-34页。 27 李晓方:《理念、激励与共享经济的敏捷治理:基于地方政府网约车监管实践的实证分析》,《中国行政管理》2019年第6期,第42-48页。 28 葛天任、裴琳娜:《高风险社会的智慧社区建设与敏捷治理变革》,《理论与改革》2020年第5期,第85-96页。 29 杨博文、孙永军:《理性赋能与向善赋权:科技伦理风险预警与敏捷治理体系的建构进路》,《科学技术哲学研究》2023年第4期,第122-128页。 30 何宇华、李霞:《生成式人工智能虚假信息治理的新挑战及应对策略——基于敏捷治理的视角》,《治理研究》2024年第4期,第142-156,160页。 31 毕文轩:《生成式人工智能的风险规制困境及其化解:以ChatGPT的规制为视角》,《比较法研究》2023年第3期,第155-172页。 32 Stone P., Brooks R. & Brynjolfsson E. et al., “Artificial Intelligence and Life in 2030: One Hundred Year Study on Artificial Intelligence,” 2016-09, https://arxiv.org/pdf/2211.06318, 2024-12-12. 33 Nabil R., “Artificial intelligence regulatory sandboxes,” Journal of Law, Economics & Policy, Vol. 19, No. 2 (2024), pp. 295-348. 34 宋科、傅晓骏:《监管沙盒的国际经验与中国应用——兼论我国“监管试点”与“监管沙盒”的异同》,《金融监管研究》2021年第9期,第100-114页。 35 Information Commissioner’s Office, “Regulatory sandbox,” https://ico.org.uk/for-organisations/advice-and-services/regulatory-sandbox/previous-participants/2020/, 2026-02-12. 36 Infocomm Media Development Authority, “First of its kind Generative AI Evaluation Sandbox for Trusted AI by AI Verify Foundation and IMDA,” 2023-10-31, https://www.imda.gov.sg/resources/press-releases-factsheets-and-speeches/press-releases/2023/generative-ai-evaluation-sandbox, 2024-12-12. 37 Datatilsynet, “Time for generative AI in the sandbox,” 2024-01-09, https://www.datatilsynet.no/en/news/aktuelle-nyheter-2023/time-for-generative-ai-in-the-sandbox/, 2024-12-12. 38 温婧:《〈北京人工智能数据训练基地监管沙盒〉成果发布 探索人工智能可控发展的创新手段》,2024年4月26日,https://new.qq.com/rain/a/20240426A068AF00,2024年12月12日。 39 Information Commission’s Office, “Regulatory sandbox final report: Yoti,” 2022-04, https://ico.org.uk/media2/migrated/4020427/yoti-sandbox-exit_report_20220522.pdf, 2024-12-12. 40 Information Commission’s Office, “Regulatory sandbox final report: Onfido,” 2020-09, https://ico.org.uk/media2/for-organisations/documents/2618551/onfido-sandbox-report.pdf, 2024-12-12. 41 Financial Conduct Authority, “Regulatory sandbox lessons learned report,” 2017-10-20, https://www.fca.org.uk/publication/research-and-data/regulatory-sandbox-lessons-learned-report.pdf, 2024-12-12. 42 Truby J., Brown R. D. & Ibrahim I. A. et al., “A sandbox approach to regulating high-risk artificial intelligence applications,” European Journal of Risk Regulation, Vol. 13, No. 2 (2022), pp. 270-294. 43 Goo J. J. & Heo J., “The impact of the regulatory sandbox on the fintech industry, with a discussion on the relation between regulatory sandboxes and open innovation,” Journal of Open Innovation: Technology, Market, and Complexity, Vol. 6, No. 2 (2020), https://doi.org/10.3390/joitmc6020043. 44 张永亮:《金融监管科技之法制化路径》,《法商研究》2019年第3期,第127-139页。 45 张苑、瞿晶晶、张晅昱:《人工智能监管沙盒:概念界定、实践经验及对我国的启示》,《世界科技研究与发展》2025年第S1期,第54-61页。 46 张红:《监管沙盒及与我国行政法体系的兼容》,《浙江学刊》2018年第1期,第77-86页。 47 黄锫:《论我国人工智能领域包容审慎监管的法治维度》,《财经法学》2025年第2期,第94-109页。 48 周辉、李延枫主编:《〈人工智能示范法〉释义》,北京:中国社会科学出版社,2024年。 49 叶宣含:《论人工智能治理的监管沙盒路径与机制(英文)》,《科技与法律(中英文)》2024年第5期,第136-148页。 50 戚聿东、刘健:《人工智能产业的包容审慎监管:理论内涵与实现路径》,《兰州大学学报(社会科学版)》2024年第4期,第154-165页。 51 Yordanova K. & Bertels N., “Regulating AI: challenges and the way forward through regulatory sandboxes,” in Antunes H. S., Freitas P. M. & Oliveira A. L. et al. (eds.), Multidisciplinary Perspectives on Artificial Intelligence and the Law, Berlin: Springer, 2024, pp. 441-456. 52 武琼:《科技向善:中国全球人工智能治理倡议的核心要义与理论价值》,《和平与发展》2024年第5期,第25-49,179-180页。 53 翟岩、李小波:《“人工智能+”视域下垂域模型的安全风险与适应性治理》,《社会科学家》2026年第3期,第96-104页。 54 Financial Conduct Authority, “Regulatory sandbox eligibility criteria,” 2022-03-27, https://www.fca.org.uk/publication/fca/fca-regulatory-sandbox-eligibility-criteria.pdf, 2024-12-12. 55 Monetary Authority of Singapore, “Fintech regulatory sandbox guidelines,” 2016-11-16, https://www.rajahtannasia.com/media/pdf/15_FinTech_RegulatorySandbox_Guidelines.pdf, 2024-12-12. 56 刘岳川:《科技创新的法律规制》,《华东政法大学学报》2023年第3期,第37-46页。 57 陈伟:《监管沙盒退出机制的国际经验及本土化研究》,《南方金融》2020年第12期,第44-53页。 58 王湘军、刘莉:《冲击与重构:社会变迁背景下我国市场监管手段探论》,《中共中央党校(国家行政学院)学报》2019年第2期,第100-109页。 59 周雷、陈佳慧、熊舒琪等:《金融强国视域下中国版“监管沙盒”的构建与完善——基于金融科技创新监管试点的应用研究》,《西南金融》2024年第4期,第60-73页。 |
|
|
|