Directing the Algorithm: Rethinking Authorship through the Improvisational Director Analogy

Kenan Başkuru LL.M.

doi:10.5771/9783689005771

Summary

This thesis examines the conditions under which human users can be recognised as the authors of images generated by generative artificial intelligence (GAI). It contends that the unpredictable and opaque nature of GAI does not automatically preclude user authorship in prompt-based image generation. Drawing an analogy with improvisational cinema, it contends that just as directors guide unpredictable performances through creative decisions, GAI users can shape outputs through iterative prompts. Based on a comparative analysis of EU, US and Turkish copyright law, the thesis proposes a process-based originality assessment and introduces the concept of 'Directed Originality'.