Psychological Interventions for Post-Traumatic Stress Disorder in Military Personnel using Artificial Intelligence

Abstract

Background:
Post-traumatic stress disorder is prevalent in active-duty military personnel and veterans, and access to trauma-focused psychotherapy is limited by workforce capacity, operational demands, and stigma. Artificial intelligence-enabled technologies may enhance delivery and engagement.
Methods:
PubMed was searched for randomized trials or comparative cohort studies of artificial intelligence–enhanced psychological interventions for post-traumatic stress disorder in active-duty personnel or veterans. Screening, data extraction, and risk-of-bias appraisal were conducted in duplicate, and findings were synthesized narratively without meta-analysis.
Results:
Eleven studies (9 randomized trials, 2 nonrandomized studies) were included, mainly immersive virtual reality exposure variants and motion-assisted reconsolidation approaches. Virtual reality–graded exposure improved response versus usual care (≥30% Clinician-Administered PTSD Scale reduction: 7/10 vs 1/9; RR 3.2), and motion-assisted therapy in treatment-resistant veterans reduced clinician-rated symptoms versus control (mean difference −9.38 points; 95% CI −17.33 to −1.44). An intensive multicomponent trauma management program reported large improvement (effect size 2.06) and 65.9% no longer meeting diagnostic criteria, while dropout in exposure-based trials reached 4-44%.
Conclusions:
Artificial intelligence-enabled interventions were associated with clinically meaningful post-traumatic stress disorder symptom reductions, but heterogeneity and adherence constraints limited certainty. Larger pragmatic trials with standardized outcomes, safety reporting, and longer follow-up are needed.