Authorship in Human-AI collaborative creation: A creative control theory perspective

Abstract

The emergence of human-AI collaborative creation (HAIC) models has provided a good opportunity to uncover the principles of authorship identification. To clarify whether humans exert control over AI-generated content (AIGC) and whether such control is sufficient to confer authorship, we propose the theory of creative control from a law and aesthetics perspective. According to this theory, a human can claim authorship when they are guided by artistic imagery thinking and manifest individual creativity throughout the entire creation process from conception to execution. In the HAIC model, the unpredictable nature of the AI black box does not impede the recognition of users’ control, as users possess the capability for artistic imagery thinking to direct the entire creation process. If their contribution meets the originality standard, they qualify as the author of the AIGC. Current prevailing views that evaluate AIGC’s originality on the basis of either the final form of expression or the users’ prompts in the initial stage overlook the dynamic nature of the creative process.