Mediating Meeting Dynamics: An Exploration of AI-Based Multimodal Feedback in Hybrid Meetings

Abstract

Hybrid meetings have become common practice in collaborative work, yet they present unique challenges in ensuring the integration of participants regardless of their physical location. Adaptive multimodal feedback, facilitated by Artificial Intelligence~(AI), can reshape and mediate hybrid settings by analysing meeting dynamics and providing timely interventions. In this exploratory study, we investigate how a multimodal visual and haptic tool mediates hybrid meetings through AI-based multimodal feedback. In two hybrid meetings within a company, we explore participants' experiences and perspectives on the role of AI and human intervention in meeting facilitation. Our findings suggest that while AI-driven mediation through subtle multimodal feedback can enhance awareness and communication in hybrid settings, participants prefer to attribute responsibility for necessary interventions, such as managing disruptions in turn-taking, to the adaptive system. Particularly in asymmetric hybrid settings, AI-generated feedback is valued due to its perceived objectivity. By examining AI's role in redistributing agency and decision-making in hybrid meetings, our study invites critical reflection on the evolving role of AI in collaborative work.