Symposium
Artificial Intelligence and Technology-based Interventions
Eduardo Bunge, Dr., Ph.D.
1791 Arastradero Road
Palo Alto University
Mountain View, California, United States
Felipe Rivera Cepeda, Mgr (he/him/his)
Professor
Universidad de Santo Tomas
Temuco, Maule, Chile
Blanca Pineda, EdD (she/her/hers)
Professor
Palo Alto University
Palo Alto, California, United States
Karin Mostovoy, MS (she/her/hers)
Ph.D candidate
Palo Alto University
Palo Alto, California, United States
Presentation: Innovating in Behavioral Parent Training: Combining Human Therapists and AI
Presenter: Eduardo Bunge, PhD
Co-Authors: Felipe Rivera Cepeda Mgr., Blanca Pineda, EdD. Karin Mostovoy, MS
Background: Around 20 % of children experience a diagnosable mental-health disorder annually, most commonly disruptive behavioral disorders. Behavioral Parent Training (BPT) programs are effective but limited by provider shortages, scheduling barriers, and costs. ParenteAI is a digital platform that was designed to extend therapists’ reach through an AI conversational agent (“Pat”) that supports parents between sessions. Because AI use in mental health is emerging, its safety, efficacy, and real-world feasibility must be evaluated.
Methods:
Two complementary studies were conducted. Study 1 assessed efficacy when ParenteAI was added to individual therapy. Study 2 examined real-world engagement, attitudes, and alliance in a retrospective cohort. In both studies, caregivers of children (ages 4–14) with disruptive behaviors completed nine BPT modules across eight live group sessions with a therapist, while Pat provided interactive guidance between sessions.
Results:
Study 1: Significant pre- to post-intervention decreases were found in child externalizing (p < .001, d = 1.00) and internalizing (p = .019, d = 0.63) symptoms, and in caregiver depression (p = .005, r = .68), anxiety (p = .002, r = .79), and stress (p = .008, d = 0.74).
Study 2: Parents reported excellent satisfaction (Net Promoter Score = 116; considered excellent), strong therapeutic alliance with Pat (M = 4.06), positive treatment attitudes (M = 4.39), 89 % module completion, and an average of 450 messages exchanged.
Conclusion:
Integrating an AI agent with human therapists produced strong engagement, satisfaction, and symptom improvements. Combined evidence from controlled and real-world settings highlights the feasibility, usability, and scalability of ParenteAI to reduce barriers to parenting programs. Findings underscore both the promise and implementation challenges of hybrid AI-assisted BPT models for everyday clinical practice.