Problematic use of generative artificial intelligence chatbots: current stage of conceptual and clinical understanding

A conceptual review that sets out what problematic use of generative chatbots is, before anyone measures it at scale. The authors gathered existing descriptions and drew the recurring symptoms from them.

Six symptoms: compulsive checking, withdrawal when access is lost, neglect of obligations, continued use despite evident harm, loss of control over time, and preferring conversations with the bot to relationships with people. The list is almost identical to the criteria for behavioural addictions — which is the main finding, not a coincidence.

For personal use it is a ready checklist to hold up against oneself. For research, a shared vocabulary without which results from different teams cannot be compared. Caveat: a conceptual review; how often the symptoms occur, and in whom, is a question for the empirical work this text prepares the ground for.