首页|生成式人工智能对著作权侵权认定接触要件的再塑造

生成式人工智能对著作权侵权认定接触要件的再塑造

扫码查看
在著作权侵权认定时,接触具有独立法律地位,且是实质性相似认定的前置性条件.接触在法律性质上应属推论而非推定,其证明也应当采用推论要求的合理可能性这一高标准.生成式人工智能介入创作后,动摇了接触规则对有限学习能力的预设,破坏了接触主体与创作主体同一性的要求,也极大地消解了接触事实验证的可能性.为应对生成式人工智能的挑战,在充分利用既有理论资源与尊重现有技术发展水平的前提下,可通过向民法规则的回归,将接触的证明转化为对行为人主观过错的证明.同时通过构建人机联合体,以集体主义取代个人主义的归责路径,化解接触主体与创作主体同一性被割裂、实际加害主体难以确认的困境.为防止使用者承担责任过重,人机联合体中使用者与生成式人工智能不宜认定为客观共同侵权,而应定性为分别侵权行为.
The Re-Shaping of the Contact Requirements for Copyright Infringement Identification by Generative Artificial Intelligence
In the identification of copyright infringement,contact has an independent legal status and is a prerequisite for substantive similarity identification.The legal nature of contact should be inference rather than presumption,and its proof should also adopt the high standard of reasonable possibility required by inference.After the intervention of generative artificial intelligence in creation,the presupposition of limited learning ability in contact requirement is shaken,the requirement of identity between access subject and creation subject is destroyed,and the possibility of contact fact verification is greatly eliminated.In response to the challenge of generative artificial intelligence,under the premise of making full use of existing theoretical resources and respecting the development level of existing technology,the proof of contact can be transformed into the proof of the subjective fault of the actor by returning to the rules of civil law.At the same time,through the construction of a human-machine consortium,the collective responsibility path replaces individualism,and resolves the dilemma that the identity of the contact subject and the creative subject is separated and the actual victimizer is difficult to confirm.In order to prevent users from taking too much responsibility,users and generative artificial intelligence in the man-machine alliance should not be identified as objective joint infringement,but should be characterized as separate infringement.

generative artificial intelligencecopyright Infringementcontactfault

王果

展开 >

天津商业大学 法学院,天津 300134

生成式人工智能 著作权侵权 接触 过错

2024

编辑之友
山西出版传媒集团有限责任公司

编辑之友

CSSCICHSSCD北大核心
影响因子:0.9
ISSN:1003-6687
年,卷(期):2024.(12)