Problematic Use of Artificial Intelligence in Young People: A Critical and Conceptual Review

abstract

The present study provides a critical and conceptual review of the problematic use of generative artificial intelligence (AI) among young people. Based on an analysis of recent literature, it examines the main labels, definitions, and theoretical approaches used to conceptualize this phenomenon, highlighting a prevailing tendency to interpret it through symptom-based models derived from behavioral addictions. The article argues that this approach may be limited and potentially pathologizing, as it tends to equate intensive AI use with forms of addiction without adequately considering the underlying psychological processes. In response, a process-oriented perspective on problematic AI use is proposed, focusing on the motivational, cognitive, and emotional processes involved in young people’s interactions with generative AI. In particular, the motivational model and the cognitive-behavioral model of problematic Internet use are integrated to explain how specific compensatory functions and self-regulation difficulties may promote dysfunctional patterns of AI use. Finally, future research directions and the practical implications of this proposal are discussed.

Keywords:

Generative artificial intelligence; Problematic generative artificial intelligence use; problematic Internet use; motivations; young people.

Received date

30/06/2026

Accepted date

14/09/2026

View article in PDF View PDF in Spanish

Authors

Mattia Pezzi - Departamento de Psicología Biológica y de la Salud, Universidad Autónoma de Madrid, España
Estibaliz Mateos-Pérez - Departamento de Psicología Social, Universidad del País Vasco, Vitoria, España
Jone Martínez-Bacaicoa - Departamento de Psicología Biológica y de la Salud, Universidad Autónoma de Madrid, España
Sophie McBain-Ritchie - Departamento de Psicología Biológica y de la Salud, Universidad Autónoma de Madrid, España
Manuel Gámez-Guadix - Departamento de Psicología Biológica y de la Salud, Universidad Autónoma de Madrid, España

DOI

https://doi.org/10.54108/10142