OBJECTIVES: This study evaluated the effectiveness of using artificial intelligence (AI) to target which patients should receive interventions compared to traditional targeting approaches to improve medication adherence. METHODS: This was a double-blinded randomized control trial (RCT) study focused on improving medication adherence to RAS antagonists, oral anti-diabetics and statins. Data was sourced from a regional health plan in North Carolina via their PBM as well as from the provider of the medication adherence interventions. Patients were randomized into three groups: Control, Traditional, and AI. Patients in the Control Group received no interventions. Patients in the Traditional Group were targeted to receive interventions based on traditional targeting methods. Patients in the AI Group were targeted based on recommendations using AI. All interventions consisted of live calls and direct mail to patients and faxes to prescribers. A Difference-in-Differences (DiD) analysis of year end adherence in 2015 compared to year end adherence in 2016 for all patients eligible in both years was used to compare each group’s impact on the population adherence. Adherence was calculated by using pharmacy claims and following CMS Star Ratings methodology. RESULTS: The study population consisted of 14,377 in the Control Group, 5,423 in the Traditional Group, and 24,527 in the AI Group. Overall, patients in the AI Group had a 6.11% increased likelihood of being adherent than the Control Group (p-value = 0.04); patients in the AI Group had a 7.8% increased likelihood of being adherent than the Traditional Group (p-value = 0.08); there was no statistically significant difference in the likelihood of being adherent for patients in the Traditional Group compared to the Control Group (p-value = 0.73). CONCLUSIONS: Utilizing AI to target interventions can increase the effectiveness of medication adherence intervention programs.