Authentic or “simulated” experience? Theoretical review and mini-case studies of generative artificial intelligence

Nowadays, users tend to misperceive the ontological nature of generative AI as quasi-social, driven by AI’s agency and humanlike behavior. However, such experience seems “simulated”—a cognitive bias that attributes authenticity to interactions that never truly occurred—and remains insufficiently explored in current experience research. To address this issue, we conduct a theoretical literature review across information systems, marketing, and service disciplines, complemented by mini case studies drawing on a qualitative analysis of 2,466 user reviews of Google’s Gemini and OpenAI’s ChatGPT. Our findings indicate that generative AI distorts affective, relational, symbolic, and cognitive experience dimensions, whereas sensorial and physical dimensions seem less affected. Theoretically, we contribute an initial framework that distinguishes authentic from “simulated” experience and treats “simulated” experience as biasing mechanism in the human mind. Our groundwork supports future studies in measuring experience while controlling this cognitive bias. Moreover, this study helps practitioners stay vigilant while designing GenAI-experience.