Design for Calibrated Trust in AI: Exploring Opportunities to Support Appropriate Mental Models When Interacting With Conversational AI
Abstract
Artificial intelligence (AI) is increasingly embedded in daily life, offering convenient support in many tasks like inspiration for text production or answering everyday questions. However, its risks often remain less visible, ranging from data security concerns to more subtle effects like growing dependence, negative impacts on creativity and reduced critical thinking. While responsible AI use is widely advocated, we argue that this responsibility cannot be achieved through education alone but must be supported through AI design. Specifically, AI systems should foster appropriate mental models in users. This workshop explores how design choices, such as reducing human-like AI design or communicating (un)certainty of responses, can support more informed and balanced use. Inviting perspectives from disciplines including computer science, psychology, design, and ethics, the workshop aims to (1) deepen understanding of failures caused by misleading mental models, (2) generate initial design strategies to address them, and (3) build a research community for future collaboration in studies, funding, initiatives, and publications.