Negotiating Acceptable Boundaries: Psychologists’ Perceptions of AI-Supported Mental Healthcare

Abstract

Artificial Intelligence (AI) and digital mental health technologies are increasingly being explored in psychological care, raising important questions about therapeutic relationships, professional responsibility, and the appropriate role of AI in clinically sensitive contexts. This paper presents findings from an exploratory qualitative study examining how psychologists working in public healthcare perceive online psychological practice and AI-supported psychological care. Using Reflexive Thematic Analysis, we analysed responses from 29 psychologists working in the public healthcare service of the Autonomous Region of Madeira, Portugal. Rather than expressing simple acceptance or rejection of AI, participants described conditional forms of acceptability shaped by organisational needs, therapeutic relationships, ethical considerations, clinical judgement, and meaningful human oversight. AI-supported systems were generally perceived as appropriate for organisational and administrative support, while core therapeutic activities were considered to require continued human involvement. The findings contribute to Human–Computer Interaction and digital mental health research by showing how psychologists negotiate acceptable boundaries for AI-supported psychological care and by identifying conditions that may inform the design of technologies that support, rather than replace, professional psychological practice.