ITeG-Ringvorlesung "The Indirect Disclosure Effect: How Disclosing Generative AI Use Impacts Creators' Collaboration with AI"

Beschreibung

In der ITeG-Ringvorlesung "Digitale Gesellschaft - eine Gestaltungsaufgabe" wird am 19. November 2025 Prof. Dr. Ekaterina Jussupow Einblicke geben zum Thema "The Indirect Disclosure Effect: How Disclosing Generative AI Use Impacts Creators' Collaboration with AI".

Ekaterina Jussupow is an Associate Professor (W2, tenure track) in Information Systems at the Technical University of Darmstadt. With a foundation in psychology, her research explores how individuals collaborate with AI systems, emphasizing the cognitive and metacognitive processes that shape these interactions. Her work has been recognized with a Best Paper Runner-Up Award from Information Systems Research (ISR), and her dissertation earned the prestigious TARGION Award. Her research has been published in leading outlets, including Information Systems Research and MIS Quarterly.

Zusammenfassung ihres Vortrags:

Text-to-image generative artificial intelligence (GenAI) systems enable creators to produce visual content through prompt-based interaction, blending human input with computational creativity. Concurrently, regulators and industry stakeholders are advancing disclosure policies that mandate transparency about GenAI involvement. While prior research has focused on how GenAI disclosure shapes audience perceptions, less is known about how anticipated disclosure affects creators’ perceptions and behaviors. Using a nested mixed-methods approach, we conducted two experiments to examine how creators interact with GenAI tools when they anticipate that judges will either learn or not learn about their GenAI use. We identify a novel indirect disclosure effect. When creators anticipate that their use of GenAI will be revealed, they reduce their creative involvement in both idea generation and execution. Specifically, they invest less effort in prompting, identify less with the result, and perceive the resulting images as a less authentic expression of themselves. This shift in the collaboration results in images perceived as more novel but less visually appealing by a judging panel. We also find some evidence that the effect of anticipating GenAI disclosure on the creative involvement is mediated by second-order beliefs, in particular, creators’ expectations that judges will perceive the image as a dishonest creative work. Interestingly, we only observe these effects on the creative process when creators know that the judging panel consists of laypeople. For an expert panel, the effects are negligibly small and statistically insignificant, hinting at a boundary condition of the novel indirect effect. Our findings underscore the socio-technical dynamics of GenAI usage and highlight how disclosure policies and the implementation of content credential technologies influence not only how creative artifacts are judged but also how they are produced. We discuss practical and societal implications for platforms and policymakers, emphasizing the need to highlight human involvement in disclosing GenAI.