Emotional AI 's Psychological Threat to Consumers: A Framework and Research Agenda

Abstract

Emotional artificial intelligence (EAI) is transforming how consumers interact with digital systems by detecting, interpreting, and responding to human emotions. While these capabilities can enhance personalization and engagement, they also create underexplored psychological risks. To address this concern, we introduce the Framework for Emotional AI Risk (FEAR) Matrix, which categorizes EAI's psychological threats along two dimensions: Pathways of Harm (mind‐driven vs. system‐driven) and Scope of Impact (individual, relational, and cumulative). Intersecting these dimensions yields six distinct threat types: Emotional Overload, Anxiety and Powerlessness, Emotional Intrusion, Dependency and Isolation, Emotional Desensitization, and Cumulative Disengagement. The framework draws on cognitive load theory, self‐determination theory, and social impact theory to explain how EAI can erode well‐being through overstimulation, loss of autonomy, blurred identity boundaries, disrupted relationships, and diminished trust. The FEAR Matrix advances prior AI–consumer research by providing a psychologically grounded, EAI‐specific structure for identifying how emotional harms emerge, differ in scope, and potentially accumulate over time. We also propose industry and policy interventions designed to enhance transparency, preserve consumer control, and mitigate these psychological risks. Together, these contributions provide researchers, marketers, and policymakers with a framework for understanding and responsibly managing the psychological consequences of EAI.