(142) Naltrexone Treatment for AI Chatbot Addiction in Adolescent With ASD and ADHD

Background &

Introduction: Adolescents with autism spectrum disorder (ASD) and attention deficit hyperactivity disorder (ADHD) face significantly elevated risk for internet addiction, with prevalence rates reaching 45% compared to 12% in the general adolescent population. Artificial intelligence (AI) chatbot platforms like Character.AI represent a particularly concerning addiction subtype, offering constant availability and personalized interactions that can become intensely reinforcing for vulnerable youth. Recent reports have documented severe consequences, including psychiatric crises and suicidal ideation in adolescents who develop intense attachments to these platforms.

The neurobiological vulnerabilities in ASD and ADHD, including impaired inhibitory control, heightened reward sensitivity, and dysregulated dopamine signaling, create a perfect storm for compulsive digital behaviors. When behavioral interventions prove insufficient, pharmacological approaches targeting the brain’s reward circuitry may offer additional support. Naltrexone, a μ-opioid receptor antagonist, has demonstrated effectiveness in reducing compulsive behaviors in gambling disorder and internet pornography addiction by modulating dopamine release in reward circuits. However, its use for AI chatbot addiction in adolescents has not been previously documented. This case describes the clinical course of a 14-year-old with severe AI chatbot addiction treated with naltrexone.

Case Description: A 14-year-old female with ASD, ADHD (combined type), major depressive disorder, and multiple anxiety disorders presented to addiction medicine in October 2025 with severe addiction to Character.AI, an AI chatbot platform. Her digital journey began early, when she received a smartphone at age 6, discovered the internet around age 9-10, and became intensely focused on Character.AI at age 12. By the time of evaluation, she was spending 4-6 hours daily on the platform, with a Chen Internet Addiction Scale score of 57-80 (moderate to severe range).

Clinical Presentation:
The consequences of her addiction were profound. She experienced two involuntary psychiatric hospitalizations (Baker Acts, Florida’s involuntary hold statute) in March and August 2025, both triggered by suicidal ideation when her parents attempted to restrict her phone access. She developed a one-inch forehead wound from compulsive skin picking related to anxiety about internet restriction. Her academic performance declined
sharply; she was failing multiple classes despite previously adequate functioning. Family relationships became strained by constant conflict over internet use, and she increasingly withdrew from real-world peer interactions, preferring her AI chatbot companions.

Treatment Course:
Despite being on optimized psychiatric medications (fluoxetine 50mg, aripiprazole 15mg, and lisdexamfetamine 40mg for ADHD) and engaged in ongoing psychotherapy, her compulsive use persisted. After discussion of off-label use, informed consent was obtained. Baseline liver function tests were normal. Naltrexone was initiated at 25mg daily for three days, then increased to 50mg daily at bedtime.
The response was notable. Within two weeks, the patient reported a striking change in her experience: she “no longer feels the large dopamine rush” when using the platform and found herself getting “bored in 5 minutes”, a dramatic shift from her previous hours-long sessions. This subjective change in reward experience aligned with naltrexone’s known mechanism of action.
By week 10, based on continued but incomplete response, the dose was increased to 100mg daily (50mg twice daily). At 14-week follow-up in January 2026, she reported feeling “more in control” of her urges, with sustained reduction in cravings.

Clinical Outcomes:
Results were encouraging across multiple domains. Most significantly, she experienced no further psychiatric crises, a stark contrast to the two hospitalizations in the six months before treatment. Depression scores improved from 22 (severe) to 12 (moderate). The forehead wound healed almost completely. Her academic performance gradually improved, and family conflict decreased. Throughout treatment, liver function tests remained normal, and she reported no adverse effects. Conclusion &

Discussion: This case offers encouraging evidence that naltrexone may help adolescents with ASD and ADHD who struggle with severe AI chatbot addiction. The patient’s description of a blunted reward response, no longer feeling the “dopamine rush” and becoming quickly bored, provides compelling phenomenological support for naltrexone’s mechanism of reducing reward salience through opioid receptor blockade. Several aspects make this case noteworthy: it represents the first documented use of naltrexone specifically for AI chatbot addiction, demonstrates the safety and tolerability of higher-dose naltrexone (100mg daily) in a medically complex adolescent, and shows meaningful clinical improvement, including the absence of further psychiatric crises. Important limitations include the inability to isolate naltrexone’s effects from concurrent psychotherapy and medications, the relatively short 14-week follow-up period, and reliance on subjective rather than objective measures of internet use. Additionally, as a single case, it cannot establish causality or predict which patients might benefit most from this approach. For clinicians treating adolescents with severe internet addiction, particularly those with neurodevelopmental vulnerabilities, naltrexone may warrant consideration as an adjunctive treatment when behavioral interventions alone are insufficient. Controlled trials are needed to establish efficacy, optimal dosing strategies, and long-term outcomes.

References:
1. Kawabe K, Horiuchi F, Ochi M, et al. Internet addiction and attention-deficit/hyperactivity disorder symptoms in adolescents with autism spectrum disorder. Res Dev Disabil.2019;89:22-28. 2.
Pan YC, Chiu YC, Lin YH. Systematic review and meta-analysis of the epidemiology of internet addiction. NeurosciBiobehav Rev. 2020;118:612-622. 3.Mazurek MO, Wenstrup C. Television, video game and social media use among children with ASD and typically developing siblings. J Autism Dev Disord. 2013;43(6):1258-1271.
4.Dichter GS, Damiano CA, Allen JA. Reward circuitry dysfunction in psychiatric and neurodevelopmental disorders and genetic syndromes: from mouse to man and back again. Neuropsychopharmacology. 2012;37(1):142-165.
5.Grant JE, Kim SW, Hartman BK. A double-blind, placebo-controlled study of the opiate antagonist naltrexone in the treatment of pathological gambling urges. J Clin Psychiatry. 2008;69(5):783-789.