The Artist is Present, but the Machine is Co-Creating- Negotiating Authorship and Agency in Generative AI-Integrated Video Art Pedagogy

Abstract

Generative artificial intelligence (Generative AI) is reshaping art and design education. In video-creation pedagogy, it has moved from an auxiliary tool to a collaborator involved in topic selection, generation, and editing. This shift raises important questions about student authorship and creative subjectivity. Existing studies focus mainly on tool effectiveness and general ethical concerns, but provide limited insight into classroom-level creative practice. This study investigates how students’ subjectivity is reconfigured within AI-supported creative workflows in the New Media Video Creation course at Xinhua University, Guangzhou, using a qualitative case study design and multi-source triangulation.

The analysis focuses on two representative tasks: Lingnan Paper-Cutting Dynamic IP Design and AI-assisted Multilingual Promotional Video for the Canton Fair. It examines problem-setting, prompt strategy, output screening, aesthetic judgment, cultural correction, workflow organization, and ethical responsibility. Three main findings emerge. First, student authorship is not dissolved. Instead, it shifts from isolated technical operation to workflow authorship, reflected in students’ ability to coordinate the creative process under AI intervention. Second, authorship is sustained through continuous negotiation, forming negotiated authorship. Students intervene at key points such as problem definition, output screening, cultural correction, dissemination adaptation, and final responsibility. Third, critical AI literacy is essential for maintaining subjectivity. Students remain active authors only when they can identify bias, explain AI participation, reflect on cultural implications, and take responsibility for expression.

The study contributes a process-oriented account of how authorship is transferred and maintained in AI-supported pedagogy. It also suggests that art education in the AI era should move beyond tool efficiency and focus on judgment, process organization, cultural correction, and responsibility.