近年,AI人工智慧的浪潮席捲全球,政府界、產業界及學術界高度關注其技術演進與工商應用。2024年被視為人工智慧應用的元年,相關產業逐漸成為未來的明星產業,社會各界認為人工智慧將改變下一階段的產業結構與日常生活。不過,人工智慧具有產業潛力,也具有諸多風險,其法制化議題逐漸在重要國家及區域受到重視。其中,歐盟的人工智慧法規被視為最進步與完備,就相關概念、術語標準、風險管理、資料治理、資料管理標準、準確性、可靠性、信任度、環境永續與網路安全有相關法律規定。我國也開始重視人工智慧法制的建構。2024年4月,立法委員就人工智慧開始提出法律草案。首先有兩部國民黨立法委員的提案,法案名稱都是人工智慧基本法草案。後續7月,代表政府行政機關的國家科學委員會也提出人工智慧法草案,但迄今仍未送至立法院成為正式提案。9月,台灣民眾黨團提出人工智慧發展及管理條例草案,成為正式提案,這是立法院黨團層次的提案。就目前觀察,國民黨立法委員及台灣民眾黨團的提案是人工智慧法制設計的較早嘗試,有必要多加重視。本文就這三部人工智慧法草案進行比較,主要關注其法制框架及規定的異同之處,並從社會學思想、行政法學管制的角度提出更廣泛的思考。
In recent years, the wave of AI artificial intelligence has swept the world, and governments, industries, and academia have paid close attention to its technological evolution and industrial and commercial applications. 2024 is regarded as the first year of the application of artificial intelligence. Related industries have gradually become star industries in the future. All walks of life believe that artificial intelligence will change the industrial structure and daily life in the next stage. However, artificial intelligence has industrial potential and many risks, and its legalization issues are gradually receiving attention in important countries and regions. Among them, the EU's artificial intelligence regulations are regarded as the most progressive and complete, and are related to related concepts, terminology standards, risk management, data governance, data management standards, accuracy, reliability, trust, environmental sustainability and network security. Legal requirements. Our country has also begun to attach importance to the construction of artificial intelligence legal system. In April 2024, legislators began to propose draft laws on artificial intelligence. First, there were two proposals from KMT legislators. The names of the bills were both the Draft Basic Law on Artificial Intelligence. In July, the National Science Council, which represents government administrative agencies, also proposed a draft artificial intelligence law, but it has not yet been sent to the Legislative Yuan to become a formal proposal. In September, the Taiwan People's Party proposed a draft bill on the development and management of artificial intelligence, which became a formal proposal. This is a proposal at the Legislative Yuan caucus level. Based on current observations, the proposals proposed by the KMT legislators and the Taiwan People's Caucus are early attempts at artificial intelligence legal design and require more attention. This article compares these three draft artificial intelligence laws, focusing on the similarities and differences in their legal frameworks and regulations, and puts forward broader thinking from the perspectives of sociological thought and administrative law control.