论我国ChatGPT类生成式人工智能的法律规制——以欧盟《人工智能法》为借鉴  

On the Legal Regulation of ChatGPT Generative Artificial Intelligence in China——Taking the EU's Artificial Intelligence Act as a Reference

作　　者：周睿[1] 刘湘琛[1] Zhou Rui;Liu Xiangchen(Hunan Normal University,Changsha,410081,Hunan)

机构地区：[1]湖南师范大学,湖南长沙410081

出　　处：《湖南警察学院学报》2024年第2期58-65,共8页Journal of Hunan Police Academy

摘　　要：以ChatGPT为代表的生成式人工智能,在语言理解与生成、知识推理等方面具备极大的潜力。然而,生成式人工智能在推动网络迅猛发展和各行业深刻变革的同时,也引发了知识产权争议、隐私及个人信息泄露等问题,且由于生成式人工智能立法缺乏高度的预见性与明确性,使得生成式人工智能在我国的应用风险加剧。所以,必须加强我国生成式人工智能的立法,从技术层面建立一个透明可信赖的算法评估标准与能够自我纠正、持续优化的AI生态系统;从法律层面借鉴、吸收欧盟对生成式人工智能立法方面的成功经验,以完善我国未来生成式人工智能发展的保护措施,为我国在人工智能领域的健康发展保驾护航。Generative artificial intelligence,represented by ChatGPT,has great potential in language understanding and generation,knowledge reasoning,and other areas.However,while promoting the rapid development of the internet and profound changes in various industries,generative artificial intelligence has also triggered such issues as intellectual property disputes,privacy,and personal information leakage.Moreover,due to the lack of high foresight and clarity in legislation for generative artificial intelligence,the application risks of generative artificial intelligence in China have intensified.Therefore,it is necessary to strengthen the legislation of generative artificial intelligence in China,establish a transparent and reliable algorithm evaluation standard and an AI ecosystem that can self-correct and continuously optimize from a technical perspective;Draw on and absorb the successful experience of the European Union in legislation on generative artificial intelligence from a legal perspective,in order to improve the protection measures for the future development of generative artificial intelligence in China and safeguard its healthy development in the field of artificial intelligence.

关 键 词：ChatGPT 人工智能 数据主体 算法评估 

分 类 号：D910[政治法律—法学]