Medically underserved and rural communities face persistent provider shortages, infrastructure constraints, and longstanding gaps in access to care. Artificial intelligence and large language models are increasingly proposed as tools to expand capacity through triage support, telemedicine, and workflow augmentation. Yet these settings also face heightened risk from biased model performance, limited clinical oversight, and fragile institutional trust. This policy memorandum distills findings from a policy-minded research review and identifies the conditions under which AI systems can be deployed to address access gaps without compounding inequities. It argues that ethical deployment in Health Professional Shortage Areas requires three linked safeguards. First, equity must be treated as a prerequisite for scale, including subgroup performance reporting, bias risk assessment, and ongoing monitoring in the intended population. Second, adoption depends on trust-building measures that preserve relational care, provide patient-facing explanations, and establish credible privacy protections and community feedback channels. Third, accountability requires clear governance rules for consent, permitted uses, and data retention, alongside liability clarity that distinguishes attended from unattended systems and specifies responsibility across clinicians, health systems, and vendors. The memo concludes that AI can support care access in underserved settings only when technical validation, culturally responsive implementation, and enforceable regulatory protections are aligned.
Anika Gundlapalli (Sat,) studied this question.