Artificial intelligence (AI) is rapidly entering health care including safety-net, rural, and underserved settings, where access, capacity, and workforce constraints are most acute.
Across the literature, AI is described as a tool that can support clinical decision-making, improve outcomes, and expand access. But those same systems also raise concerns about bias, data limitations, and uneven impacts on underserved populations.
This Digging Deeper series draws from a systematic review of peer-reviewed research to explore both sides of that tension: what AI could make possible, and what risks must be addressed to ensure it does not reinforce existing inequities.
Literature Review Prepared by IC² Institute | The University of Texas at Austin
Article Authored by Briana Martin | EHF
Start with the real problem: capacity constraints
When you step back from the technical details, one clear theme emerges from the literature: the strongest case for AI isn’t about innovation; it’s about helping health systems do more with limited time, staff, and resources.
Safety-net, rural, and underserved health systems are often working with limited staff, limited time, and limited access to specialists. The question the literature keeps returning to is simple: where could AI help fill those gaps?
When clinicians lack time: decision support and faster interpretation
One of the most common use cases is decision support. The literature describes AI systems that can analyze large volumes of patient data and help clinicians make faster, more accurate decisions.
In practice, that might mean identifying patterns in patient records, flagging potential risks, or helping interpret imaging results. In areas like radiology, for example, AI tools can detect patterns that are difficult to see with the naked eye and support more targeted care pathways.
In settings where providers are already balancing heavy workloads, this kind of support isn’t just about efficiency; it can feel like added clinical capacity.
When specialists are scarce: telemedicine and remote monitoring
Access looks different in rural and underserved settings, where specialty care may require long travel or long wait times.
The literature points to AI-supported telemedicine and remote monitoring as ways to bring care closer to patients. These tools can support ongoing monitoring, help clinicians track conditions over time, and extend the reach of existing providers.
It’s not that AI replaces specialists, but in these contexts, it can help bridge gaps where in-person care isn’t immediately available.
When care gets complex: helping make sense of information
Health care decisions are rarely simple, especially for patients with multiple health conditions or complicated treatment plans.
The literature highlights how AI can assist with this complexity and support treatment planning, identify drug interactions, and improve diagnostic accuracy. More recently, there’s growing attention on generative AI tools that can synthesize information and even suggest potential diagnoses based on symptoms.
Non-medical drivers of health: using AI to surface NMDOH and strengthen navigation
Another thread that runs through the literature is the role of social and economic factors in shaping health outcomes.
AI is increasingly being explored as a tool to identify and predict non-medical drivers of health (NMDOH), things like housing stability, income, or access to transportation. In some cases, these tools are being used to strengthen patient navigation programs by helping identify barriers and tailor support.
Notably, AI isn’t positioned as a standalone solution. Instead, it’s described as a way to surface needs earlier and support the navigators and care teams.
When communication breaks down: language access
Communication is another area where the literature points to potential gains.
Language barriers can make it harder for patients to understand care plans or communicate with providers, which can directly affect outcomes. AI-based translation tools are described as one way to support more consistent communication, especially when interpreter resources are limited or hard to coordinate.
In safety-net and underserved settings where resources are often stretched, this kind of support can help reduce delays and improve clarity during care.
The promises of AI depend on implementation
Even in the sections focused on promise, the literature is clear on one point: none of these benefits are guaranteed.
Each opportunity, whether it’s decision support, improved access, or better visibility into social needs, depends on how tools are designed, what data they’re trained on, and how they’re implemented in real-world settings.
That brings us to the next blog in the series. If these are the points where AI could help, what happens when things go wrong?
Overview and methodology
To make sense of this fast-moving space, Episcopal Health Foundation partnered with the IC² Institute at The University of Texas at Austin to conduct a structured, systematic review of the literature.
Using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework, the research team searched major databases, focusing on artificial intelligence in health care and its implications for equity, particularly in safety-net, rural, and underserved settings.
From a large pool of studies, the team narrowed in on literature reviews and supporting sources that directly addressed how AI is being used, where it shows promise, and where risks are most likely to emerge for underserved populations.
In total, the review draws on dozens of sources to synthesize key themes offering a grounded, evidence-based look at both the opportunities and the challenges of AI in health care.
What comes next in this series
Stay tuned for the next two posts:
- How AI Deepens Inequity: Where the Risks Show Up in Practice
- What It Takes to Use AI Responsibly: Guardrails for Safety-Net, Rural, and Underserved Care
ATTRIBUTION & SOURCES
This post draws from Literature Review: AI in Safety-Net Health Care, prepared by the IC² Institute at The University of Texas at Austin.
More from Digging Deeper
Where AI Could Help: Filling the Gaps that Safety-Net, Rural, and Underserved Systems Face Every Day
Artificial intelligence (AI) is rapidly entering health care including safety-net, rural, and underserved settings, where access, capacity, and workforce constraints are most acute. Across the literature, AI is described as a tool that can support clinical decision-making, improve outcomes, and expand access. But those same systems also raise concerns about bias, data limitations, and uneven
The opportunity to advance diabetes prevention through Texas Medicaid and the Rural Health Transformation Fund
EHF’s Shao-Chee Sim examines how new state and federal policy opportunities could help Texas expand and sustain diabetes prevention.
AI in Safety-Net, Rural, and Underserved Health Care: Why the Stakes Are Different
Artificial intelligence (AI) is rapidly entering health care including safety-net, rural, and underserved settings, where access, capacity, and workforce constraints are most acute. Across the literature, AI is described as a tool that can support clinical decision-making, improve outcomes, and expand access. But those same systems also raise concerns about bias, data limitations, and uneven