Keywords

Generative AI, Content Creation, Task Type, AI Delegation, AI Suggestion Integration, Creation Agency

Abstract

Artificial intelligence (AI) agents, particularly in the form of generative AI, are becoming more useful in creating content and can meaningfully serve as collaborative partners to human creators. Thus, it becomes increasingly important to develop a baseline understanding of when and how people choose to interact with AI in creating content. While information systems (IS) research has identified many important factors leading to adoption and use of traditional information technologies, the factors influencing decisions to incorporate or reject AI assistance remain underexplored. This study develops an empirical model of human-AI collaboration that integrates a well-established framework of human creative work with the recently developed concept of IS delegation. We examine how factors of humans’ creation agency (enduring involvement, knowledge self-efficacy, ambiguity tolerance, and identity threat) and different types of tasks (creative versus planning) influence people’s decisions to delegate tasks to AI and integrate AI-generated suggestions. We tested our proposed model using an experiment involving over 300 participants completing AI-assisted writing tasks on a purpose-built online platform that tracked users’ microbehaviors. Our preliminary results confirm that certain aspects of creation agency inform humans’ AI delegation and integration decisions, though the effects are sometimes contingent upon the type of task at hand. We hope this research-in-progress lays the groundwork for future research that enhances productive but healthy human-AI collaboration.

Download

Share

COinS