Rethinking computational models of addiction: toward context-sensitive and psychologically grounded frameworks

Although addiction is a complex and contextually embedded disorder that extends beyond individual pathology and neurobiological dysfunction, prevailing computational and clinical models often reduce addiction to a chronic brain disease. While such frameworks have shaped dominant approaches to treatment and theory, they remain poorly aligned with the lived experience and behavioral phenomena of addiction, ignoring its psychological, social, and systemic dimensions. This paper examines the limitations of various disease and compulsion models both critically and in-depth, highlighting their empirical and conceptual shortcomings. In doing so, it argues for the development of context-sensitive and psychologically grounded computational models, ones capable of capturing the nuanced realities of addiction and informing more effective, personalized interventions.