생성형 인공지능과 창작 문화: 권리 귀속과 법적 보호의 통합적 접근

Purpose: This study examines the attribution of rights and the establishment of a legal protection framework for works created by generative artificial intelligence (AI). Focusing on copyright, patent law, and publicity rights, it explores the legal challenges and gaps that arise when AI-generated content intersects with human-centered intellectual property regimes. Methods: A qualitative literature review and comparative legal analysis were conducted. Key judicial cases (Thaler v. Perlmutter, Zarya of the Dawn, Getty Images v. Stability AI), international legislative developments (EU AI Act, EPO’s DABUS decision, Japan’s AI-assisted works framework), and Korean policy documents (KRW, MCST) were analyzed. The research proceeded in four stages: examination of AI’s technical structure, assessment of copyright risks in training data, evaluation of authorship and inventorship requirements, and comparison of national legal approaches. Results: Findings indicate that AI-only works remain unprotected under current systems, as copyright and patent law require human authorship or inventorship. Nevertheless, Japan’s flexible recognition of “AI-assisted works,” the EU’s emphasis on transparency, and U.S. case law reaffirming human authorship highlight emerging approaches. Conclusion: A new paradigm is required to balance innovation with rights protection. Policy implications include establishing a legal category for AI-assisted works, strengthening data governance, codifying publicity rights in Korea, and promoting international harmonization.