Paper Type

ERF

Abstract

This study examines the psychological mechanisms underlying individuals’ adherence to AI-generated health recommendations. Drawing on communication theory and trust-in-automation research, we develop a conceptual model explaining how message-level characteristics influence trust in AI and subsequent behavioral adherence. Specifically, we propose that message grounding - defined as the extent to which AI recommendations incorporate user-specific information - enhances trust in AI. We further argue that perceived AI consistency strengthens this relationship by reinforcing system predictability. AI self-efficacy is theorized as an additional antecedent of trust, reflecting individuals’ confidence in their ability to effectively engage with AI systems. Trust in AI is expected to promote adherence to AI feedback, while perceived risk moderates this relationship by shaping users’ willingness to accept vulnerability. A 2×2 experimental design is proposed to test the model. The study advances understanding of behavioral compliance with AI health advice and informs the design of trustworthy generative AI systems.

Paper Number

1490

Comments

AI SYSTEM

Download

Share

COinS
 
Aug 15th, 12:00 AM

Evaluating Individual Adherence to AI Health Recommendations

This study examines the psychological mechanisms underlying individuals’ adherence to AI-generated health recommendations. Drawing on communication theory and trust-in-automation research, we develop a conceptual model explaining how message-level characteristics influence trust in AI and subsequent behavioral adherence. Specifically, we propose that message grounding - defined as the extent to which AI recommendations incorporate user-specific information - enhances trust in AI. We further argue that perceived AI consistency strengthens this relationship by reinforcing system predictability. AI self-efficacy is theorized as an additional antecedent of trust, reflecting individuals’ confidence in their ability to effectively engage with AI systems. Trust in AI is expected to promote adherence to AI feedback, while perceived risk moderates this relationship by shaping users’ willingness to accept vulnerability. A 2×2 experimental design is proposed to test the model. The study advances understanding of behavioral compliance with AI health advice and informs the design of trustworthy generative AI systems.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.