Abstract
Multimodal AI systems promise more natural, expressive, and creatively rich interactions, yet they also complicate how users convey intent and maintain control. As these systems rapidly enter everyday practice, agency becomes shared between humans and machines, raising urgent questions about: fairness, trustworthiness, responsibility, and accountability; calibrated trust and appropriate reliance; creative authorship and co-creation; meaningful human-AI decision-making. The goal of this workshop is to clarify the concept of distributed agency and explore its implications for human and artificial creativity. Through presentations and hands-on activities, participants will collectively develop human-centered design pathways for multimodal AI interfaces.