|
|
|
| The Practical Challenges and Countermeasures of Patent System in the Era of Artificial Intelligence |
| Deng Peng,Li Fang,Li Mingjing |
| (Intellectual Property Development Research Center, China National Intellectual Property Administration, Beijing 100000, China) |
|
|
|
|
Abstract Nowadays with the increasing complexity, the autonomous decision-making abilities and deep applications in a wider range of fields, AI systems have posed significant challenges to the current legal and regulatory systems. Particularly, in the field of intellectual property, the entire ecosystem is facing unprecedented challenges, from technical solutions to the review of inventiveness, and from assisted decision-making to autonomous inventions.#br#In this article, patent systems of artificial intelligence in major countries are sorted out, two characteristics of positive incentive and reverse constraint are summarized. On the one hand, the object of the patent protection is definite and the uniqueness in the internal structure design of the system is pursued. On the other hand, since artificial intelligence system has prominent autonomy, monopoly and invisibility, the public is seriously concerned with a series of risks and issues, such as security, privacy, integrity, dignity, autonomy and data rights. The sound development of AI in a secure and trusted system has become the focus of attention. To this end, the guidance of "justice" is important, and the construction of collaborative governance system and security barrier should be explored to eliminate the potential harms.#br#This study further analyzes the practical challenges of the current patent system from the perspective of technology and application of artificial intelligence. From the technical perspective, the " black box of algorithm" leads to complexity of the novelty, creativity and practicability of patents. To be more specific, the algorithm black box makes it extremely difficult to judge and deduce the technical scheme. Especially when the artificial intelligence system is applied to specific field scenes to perform specific tasks, it is almost impossible to accurately understand all the thinking processes and details of the machine, and even the inventor may be confused about the algorithm; the current algorithm is difficult to ensure fairness in judging the "creativity standard"; the high unpredictability in the algorithm decision-making process is likely to lead to the lack of repeatability and practicality of the invention, which may hinder the application and progress of technology. Model defects leads to the defects in applicability and robustness of the patent technology solutions. The data crisis brings the problems in openness, transparency, interpretability and traceability. From the perspective of application, the identification and supervision of the products of artificial intelligence, the impacts of inventors of artificial intelligence, the monopoly of patent technologies and markets, and the security and credibility of technology have all posed great challenges to the current patent system.#br#The study focuses on the practice of patent system, and puts forward four suggestions according to the above challenges. First, the patent disclosure rules of architecture system of algorithm, objective function and data should be improved to promote the innovation, application and flow of artificial intelligence technology. Second, the top-level design should be improved, and the relevant patent application process and review means should be optimized, thus meeting the urgent needs of innovation protection. Third, the identification system of artificial intelligence products should be established to provide basic guarantee for effective identification and management. Fourth, the linkage role of the patent systems should come into play to help build long-term risk prevention and control mechanisms dominated by technical regulation and legal control.#br#By analyzing the impacts and challenges of the patent system in the era of artificial intelligence, this paper puts forward specific suggestions on the optimization and improvement of the patent system related to artificial intelligence in China. With the continuous optimization and improvement of the intellectual property system represented by patents, the deep integration of artificial intelligence technology and industry will be promoted, and the high-quality development of the whole society will be boosted.#br#
|
|
Received: 15 May 2021
|
|
|
|
|
|
| [1] Data processing-artificial intelligence[EB/OL].[2021-03-05]. https://www.uspto.gov/web/patents/classification/uspc706/defs706.htm.[2] USPTO.Public views on artificial intelligence and intellectual property policy[EB/OL].[2020-10-07].https://www.uspto.gov/sites/default/files/documents/USPTO_AI-Report_2020-10-07.pdf.[3] 国家知识产权局关于修改《专利审查指南》的公告（第343号）[EB/OL].[2021-02-10]. https://www.cnipa.gov.cn/art/2019/12/31/art_2073_143003.html.[4] 吴汉东.人工智能生成发明的专利法之问[J].当代法学,2019,33(4):24-38.[5] Civil law rules on robotics（2015/2103（INL））[EB/OL].[2020-10-15]. www.europarl.europa.eu.[6] 王瀚. 欧美人工智能专利保护比较研究[J]. 华东理工大学学报(社会科学版), 2018,33(1):96-101.[7] 崔鹏,邓柯.安全可信智能的可能技术路径[J].中国计算机学会通讯,2020,20(11):23.[8] 何积丰. 安全可信人工智能[J]. 信息安全与通信保密, 2019,310(10):7-10.[9] 徐凤.人工智能算法黑箱的法律规制[J].东方法学，2019,12(6):78.[10] 李婕.人工智能中的算法与法治公正[N].人民法院报，2018-05-23（2）.[11] CHALMERS D J, FRENCH R M, HOFSTADTER D R. High-level perception, representation, and analogy[J]. Journal of Experimental & Theoretical Artificial Intelligence, 1992（5）.[12] 孜里米拉-艾尼瓦尔,姚叶.人工智能技术对专利制度的挑战与应对[J].电子知识产权,2020,30(4):56.[13] 李莉.人工智能专利保护[J].电子技术与软件工程,2019,8(12):248.[14] 邓建志, 程智婷. 人工智能对专利保护制度的挑战与应对[J]. 南昌大学学报(人文社会科学版), 2019,50(2):21-23.[15] 吴汉东, 张平, 张晓津.人工智能对知识产权法律保护的挑战[J].中国法律评论,2018,5(2):1-24.[16] 杜园春，张若白，渠性怡.51.3%受访者遭遇过大数据“杀熟”[N].中国青年报，2018-03-15(2).[17] 陈根.人工智能时代，技术不中立[EB/OL].[2021-01-11]. https://m.thepaper.cn/baijiahao_10739336.[18] 吴汉东.人工智能时代的制度安排与法律规制[J].法律科学（西北政法大学学报），2017,35(5):128-136.[19] 邓鹏. 数字经济时代知识产权的机遇与挑战[J]. 中国发明与专利, 2020,20(9):12-17.[20] USPTO.Inventorship limited to natural persons[EB/OL].[2020-10-05]. https://www.uspto.gov/sites/default/files/documents/16524350_22apr2020.pdf?utm_campaign=subscriptioncenter&utm_content=&utm_medium=email&utm_name=&utm_source=govdelivery&utm_term=.[21] 李想.人工智能参与发明的授权问题探究[J].科技进步与对策，2020,37(15):144.[22] 马忠法,彭亚媛,张驰.与人工智能相关的主要知识产权法律问题[J].武陵学刊,2019,44(1):52.[23] 张乃根. 论人权与知识产权的关系[C].中国人权研究会;中国人权发展基金会, 2012.[24] ROBERT JEHAN,RYAN ABBOTT.Should an AI system be credited as an inventor[EB/OL].[2020-08-24]. http://artificialinventor.com/should-an-ai-system-be-credited-as-an-inventor-robert-jehan/.[25] 黄晋.人工智能会不会促进垄断[N].中国社会科学报，2018-10-10（3）.[26] SUKHAYL NIYAZOV.AI-powered monopolies and the new world order[EB/OL].[2020-06-27]. https://towardsdatascience.com/ai-powered-monopolies-and-the-new- world-order-1c56cfc76e7d.[27] 李开复.AI·未来[M].杭州：浙江人民出版社，2018.[28] 刘鑫. 人工智能生成技术方案的专利法规制——理论争议,实践难题与法律对策[J].法律科学(西北政法大学学报), 2019,37(5):82. |
|
|
|
