The integration of artificial intelligence (AI) at the Department of Veterans Affairs (VA) is accelerating as the agency phases out older programs. This strategic realignment focuses on key priorities like suicide prevention and updating electronic health records (EHRs). VA leaders spoke on this development at the Potomac Officers Club’s 2025 Healthcare Summit, emphasizing AI's role in enhancing patient safety and clinical processes.
One of the significant components of this effort is the Recovery Engagement and Coordination for Health–Veteran Enhanced Treatment (REACH VET) program, initially launched in 2017. This predictive model assists in identifying veterans at high risk of suicide, facilitating proactive outreach. An improved version was introduced last year, incorporating more risk factors, including military sexual trauma, to refine its predictive accuracy. The VA is also initiating a project to analyze data from the Veterans Crisis Line, focusing on linguistic and acoustic indicators to better identify urgent suicide risks.
Despite these advancements, oversight regarding AI implementation remains a concern. A recent report from the VA Office of Inspector General pointed out deficiencies in managing risks associated with generative AI in clinical environments. The VA responded by asserting that AI tools will function as supportive resources, with clinical decisions to be made solely by qualified personnel. As the department prepares to restart federal EHR rollouts at 13 facilities, early AI pilot programs are anticipated to guide the integration of AI into the new EHR systems.