디자인보호법상 생성형 인공지능 디자인의 창작자성 판단기준에 관한 연구

This study examines the necessity of re-establishing the criteria for determining authorship under design protection law, premised on an industrial design environment that has shifted toward human–GAI collaborative workflows. From a comparative legal perspective, the study analyzes the normative purposes and evaluative frameworks underlying the authorship standards articulated in: (1) the USPTO’s doctrine of “significant contribution,” including the factors specified in the Pannu guidance; (2) the U.S. Copyright Office’s principle of “human authorship” as an essential condition for copyrightability; and (3) the concept of “creative contribution” applied by the Korean Ministry of Culture, Sports and Tourism and the Korea Copyright Commission in the context of works incorporating GAI-generated components. These frameworks share the foundational principle that the mere use of GAI does not negate human authorship, and that authorship must ultimately originate from a natural person. A normative convergence is also observed in the central role assigned to evidentiary substantiation of contribution (attribution) when assessing eligibility for authorial or inventive status. Building on these insights, this study aims to derive refined criteria for assessing design authorship under design protection law. As a preliminary matter, the concept of creativity and the doctrine of non-obviousness—presumed by design law—are normatively and functionally distinct from the originality standard in copyright law, reflecting fundamentally different regulatory objectives. Consequently, attribution models and the architecture of disclosure obligations must inevitably diverge across legal regimes, as the mere fact of utilizing GAI cannot carry identical normative weight under systems governed by different statutory purposes and evaluative axes. This study further distinguishes between (i) the process-oriented requirement of Significant Contribution during human use of generative AI tools, and (ii) the outcome-oriented standard of Substantial Contribution at the stage of conceiving or concretizing the dominant features of a completed design. Significant Contribution functions as a procedural threshold for recognizing a natural person as a design author when GAI is employed as a tool. It evaluates the quality and decisional agency of human intervention in the design process itself, operating as an entry-level criterion for authorship eligibility. By contrast, Substantial Contribution functions as a conclusive, outcome-focused standard that identifies the origin of the design’s dominant, rule-defining characteristics and traces those characteristics to their true conceptual source, thereby determining the legitimate rights holder entitled to registration. Accordingly, the study proposes that these two standards operate on independent, analytically distinct axes: the former governing authorship eligibility in AI-assisted workflows, and the latter governing rights attribution for a finished design. It further argues that achieving normative coherence in human–GAI co-creative design requires a staged, complementary assessment structure in which process-based and outcome-based contribution review are considered sequentially and holistically, in order to ensure both lawful recognition of human authorship and equitable vesting of registration rights.