What Counts as Proactive? Rethinking Proactivity in Conversational Agents

Human-Centred Explainable Conversational AI: Explainability, Persona Adaptation, and Entrepreneurial Agency

The Illusion of Progress: Dialogue-Derived Signals for Proactive Intervention Timing in Student–GenAI Interaction

Why This and Not That? A Collaborative Reflection Approach for Understanding Thought Coverage in Decision Making Support Dialog

This work pauses a reflection-support agent when it would normally redirect the conversation, surface its observation, and asks users to interpret the pattern and decide how to proceed, and derives a taxonomy of nine interpretation categories and shows that similar reflective states can call for substantially different follow-up actions.

Human as Relation: How Do We Shape Our Relationship with AI from a Dialogue Perspective?

This study examines how human-AI relationships can be reinterpreted through a dialogue-centred perspective within the framework of Human-Computer Interaction (HCI). Building on Bødker’s three waves of HCI, we trace the evolving role of the human from a cognitive “factor” to an intentional “actor,” and ultimately to a relational agent embedded in socio-technical contexts. While early approaches focused on usability and error reduction, later paradigms acknowledged emotions, goals, and lived experiences. In the third wave, interaction is recognised as emergent, situated, and relational, shaped through ongoing negotiation between humans and non-human agents.We propose that dialogue offers a productive lens to understand and design these relationships. Dialogue here refers not only to language but also to how meaning, form, behaviour, and responsiveness are cocreated during interaction. To analyse how such dialogic relations develop, the study adopts a motivation-based analytical approach, drawing on Maslow’s hierarchy of needs as a framework for understanding how unmet needs activate engagement and shape user expectations. Through a qualitative analysis of selected human – AI interaction examples, we explore which needs are addressed and what relational outcomes emerge. By combining theoretical insight with motivational analysis, this work aims to inform future design practices that support more meaningful human-AI relations. Methodologically, the study employs a qualitative, theory-driven analysis of selected human–AI interaction cases examined through Maslow’s hierarchy of needs to identify how different motivational layers shape dialogic relationships.

Trust in Robots: Aligning with Misaligned actions

This article examines cases from experiments on human–robot interaction in an educational context. Drawing on ethnomethodology and conversation analysis, I analyze interactions structured around the sequential organization of [question]–[answer]–[evaluation] to explore the ascribability of “trust” as a routine foundation of sociality. The analysis focuses on a recurring phenomenon in which participants deal with audio input issues and respond to the robot's misaligned actions by laughing at the robot's behavior and repeating their response. I observe how, in human–robot interaction, robots enact actions programmed to resemble human interactional competence and how the participants align with the misaligned actions while managing the activities that the robots initiate. Moreover, I demonstrate that participants treat the fragility of this resemblance as a basis for determining the robot's trust‐ability.

Clarify User Expertise: Towards Proactive Conversational Agents Tailoring Responses to User Proficiency

In the context of information seeking, conversational agents are undergoing an evolution from reactive tools to proactive, personalized assistants. A critical aspect of this evolution is the ability to tailor strategic interactions to a user's unique needs and expectations. Unlike existing studies that focus on proactively clarifying query ambiguities, we center on clarifying the user's expertise in order to tailor responses for better user comprehension. We find that existing agents struggle to determine user expertise from queries alone, a limitation that prevents them from dynamically adapting their responses. To address this gap, we introduce PASSING to empower the agent to proactively clarify a user's expertise through targeted inquiries. This is achieved by our What-to-ask and How-to-ask strategies, induced by LLM self-play. Our extensive experiments also show our superiority. We believe that PASSING represents a crucial step towards creating more human-centric conversational agents.