首页|论人工智能生成内容的可版权性:以用户的独创性表达为视角

论人工智能生成内容的可版权性:以用户的独创性表达为视角

扫码查看
AIGC既包含来自机器的选择,也包含来自人类用户的选择.当后者满足独创性表达要求时,AIGC便足以被认定为作品.在判断用户的独创性表达时,裁判者应当关心人贡献了什么,而不是工具贡献了什么;应当关心人贡献了什么,而不是人没有贡献什么;应当关心人贡献的实质,而非纠结于其形式.在以上原则的指导下,既有的作品构成要件规则足以支持相当一部分AIGC获得作品资格.AIGC的所谓"随机性",只是发生在用户指定范围内的"随机",并不妨碍用户控制达到著作权法的要求.作为文本的提示词,具有转化为包括视觉表达在内的其他类型表达的可能性.若以摄影为参照系,AI与传统工具的比较不仅不能导致对作品资格的否定,反而能够凸显"工具贡献大并不意味着人的贡献不足".著作权法的"宽进宽出"结构提示我们,将用户做出独创性表达的AIGC纳入著作权法图式是在认知层面最为经济的利益平衡分析框架.考虑到独立创作例外和版权救济手段的灵活性,承认AIGC获得作品资格的可能性并不会过度妨碍公众自由.
Artificial Intelligence Generated Content(AIGC)reflects choices made by both users and machines.When choices made by users satisfy the threshold of copyrightability,AIGC should be acknowledged as a work.Decision-makers should focus on human users'contribution rather than machines'contribution,on what human users contribute rather than what they do not,and on the substance rather than the forms of the contribution.Under these guidelines,a significant portion of AIGC is likely to be copyrightable works.The perceived"randomness"of AIGC is confined to areas designated by users,making the content non-random from the user's perspective and not hindering the copyrightability of AIGC in a lump-sum way.Notably,prompts in textual content could potentially be embodied in visual arts.Examining the history of copyright and photography reveals that machine contributions do not necessarily diminish human authorship to the point where users are not eligible to be authors.Recognizing the possibility of AIGC being considered as works aligns with the most cognitively efficient way to balance the interests of creators and users.Copyright law provides sufficient room for the public to act freely,including the rule on independent creation and flexibility on remedies for infringement.

AIGCoriginalityidea-expression dichotomyfixationcognitive economy

蒋舸

展开 >

清华大学法学院

人工智能生成内容 独创性 思想/表达二分法 固定性 认知经济性

北京市社会科学基金

20FXB008

2024

知识产权
中国知识产权研究会

知识产权

CSSCICHSSCD北大核心
影响因子:1.115
ISSN:1003-0476
年,卷(期):2024.(1)
  • 1
  • 73