Persuasive AI

Persuasive Artificial Intelligence (AI) explores the design of "personalized adherence companions" powered by generative AI. In professional counseling and advisory settings, a significant challenge is the "adherence gap", i.e., the disconnect between receiving expert advice and consistently acting on it. Traditional persuasive systems often rely on static, "one-size-fits-all" designs and operate as standalone applications that neglect the central role of the human expert. Persuasive AI overcomes these limitations by acting as an intelligent digital bridge between the expert and the client. It leverages the dynamic content generation and deep personalization capabilities of generative AI to deliver context-aware support, helping individuals achieve their goals while effectively supporting and scaling the essential work of human experts.

In healthcare, adherence to treatment is essential for managing chronic diseases that require long-term lifestyle adjustments regarding dietary and physical habits. The Digital Companion project utilizes chatbots to bridge the gap between medical consultations, a period where patients are often left to their own devices. Due to their high availability and scalability, these conversational agents can deliver engaging, personalized messages that support patients in their daily routines. This approach aims to enhance overall treatment adherence by maintaining a continuous, supportive connection that acts as a digital ally between the patient and their healthcare professional.

This research area explores the integration of persuasive systems into mental health care, with a particular focus on supporting therapies such as cognitive-behavioral therapy (CBT). We are developing a personalized conversational companion designed to assist patients during the periods between active psychotherapeutic sessions. By providing a responsive web-based application, patients receive reminders, engage in guided exercises, and track their progress, which boosts motivation and adherence. Furthermore, the system offers therapists a transparent overview of a patient's efforts between sessions, enabling them to prepare more effectively for the next appointment.

Procrastination represents a common adherence gap in academic and professional development, frequently leading to delayed tasks, heightened stress levels, and suboptimal learning outcomes. To address this self-regulatory failure, this project develops a digital study buddy that brings awareness to users during moments of procrastination. The platform utilizes various persuasive strategies to offer personalized, interactive support. Ultimately, the project aims to generate actionable design principles for AI systems that empower students to mitigate procrastination and promote productive, long-term study habits.

• Wenyuan "Rafael" Wu