Reducing the cost of care means making every interaction count.
Whether it’s helping a member reach the right specialist, catching a care gap before it becomes a crisis, or giving a clinician better information at the point of care, care coordination has always been one of healthcare’s most powerful levers for improving outcomes. AI is now reshaping how health plans pull that lever.
At the Emids Healthcare Summit 2026 in Nashville, I joined a roundtable of healthcare leaders to talk through how their organizations are rethinking care coordination for an AI-enabled future. The conversation kept coming back to one thing: removing the barriers that keep members from getting the care they need, and making that care more connected, more proactive, and more personal.
Five themes surfaced repeatedly.
1. Put the member at the center
Health plans organize around products, lines of business, benefit structures, and workflows. Members don’t.
For a member, healthcare is one continuous experience. They expect it to feel coordinated, regardless of who’s providing the care, who’s paying for it, or which department is managing their benefits.
One participant put it plainly: plans need to stop tracking covered lives by line of business and start recognizing that every member, regardless of age or plan type, has a service they need and a goal behind it.
That reframing changed how the room talked about who is ready for AI. One participant described a home-based care program built around the highest-risk, highest-need members, the population plans most often assume will resist new technology. The organization expected resistance and got the opposite: those members received AI-enabled services with enthusiasm and became some of the most receptive people in the program. That same team is now distributing wearable rings to pilot members, pairing passive health data with in-home care coordination, a concrete step toward meeting people where they already are.
The same principle showed up in how plans reach members, not just how they treat them. Not everyone wants a phone call. Some want a text, some want a portal message, some want nothing until they ask for it. Several leaders pointed to matching the channel to the member as one of the more overlooked wins available to plans today.
Not every plan’s member focus is philosophical, and one leader was candid about that. Managed Medicaid plans prioritize members partly because government mandates require it. That’s a fair point to sit with: regulation can be the reason an organization gets member-centricity right, and that’s still a win worth taking.
Don’t assume who will embrace new ways of getting care. Start by understanding what the member needs.
2. Better context leads to better decisions
Clinical data is necessary. It’s rarely sufficient. Care teams need visibility into social determinants of health, behavioral patterns, and care history alongside the clinical record. Transportation, food access, housing, and utilities can matter more to a member’s health than the diagnosis in the chart and plans often only learn about them when a case manager happens to ask the right question.
The upside of closing that gap showed up in the numbers.
In one home-based program, giving clinicians visibility into social and environmental barriers alongside the clinical chart pushed home-setting utilization up 40%. Those members were finally getting care they needed and hadn’t been getting.
Context cuts the other way too. One participant noted that cost per patient can vary sharply from one provider to the next for the exact same service, and that for managing cost, who performs the service matters more than who made the referral.
The goal is getting the right information to the right person the moment they need it, not simply more data.
3. AI works best when it supports human judgment
AI came up constantly in the room, but never as a replacement for the clinician, care coordinators, or case managers.
Leaders described it instead as an accelerator: summarizing calls, consolidating clinical information, flagging care gaps, and recommending next-best actions so people spend less time on paperwork and more time on the interactions that matter. One participant described using AI to process inbound prior authorization calls for OB services, cutting the manual work of pulling out what’s needed from each call.
The room was equally clear about where AI’s authority stops. In practice, AI could approve a recommendation but not deny one, a limit tied directly to scope-of-practice and credentialing rules, not just caution. Recommendations still must be evidence-based, explainable, and traceable back to their source, and a human stays accountable for anything with clinical, financial, or regulatory weight.
One leader summed up the shift plainly: AI made it easier to trust the judgment they already had, because the reasoning behind a recommendation is now visible instead of buried.
The real opportunity is AI processing information at a scale no team could manage alone, paired with the judgment only an experienced clinician brings.
4. Technology alone won’t fix care coordination
The most candid part of the discussion had little to do with the tools themselves. It centered on everything around them: legacy systems, organizational silos, budget limits, and workforce readiness, all of which slow progress more than the technology does.
Plans differ from each other too. A national plan with a large self-funded book controls its benefits differently than a local, fully insured plan does, and that difference shapes how far and how fast an AI initiative can realistically scale.
AI can’t stay boxed inside an innovation team. Several leaders pointed to education and hands-on training as the real driver of adoption, giving employees across the organization a concrete sense of how AI fits the work they already do.
Establishing baseline metrics before rolling out AI is what lets a plan prove impact on quality, efficiency, and member outcomes later, rather than arguing about it after the fact.
Budget and legacy thinking are real constraints, and none of them require waiting to start.
5. Move with purpose
Healthcare has always approached innovation cautiously, for good reason. The conversation pushed past caution into a sharper question: when does caution become inertia?
One participant asked the room directly whether failing to adopt AI is now an existential risk for the plans that don’t. Another was more direct still: at the scale plans operate today, some problems simply can’t be solved without it.
Moving quickly doesn’t mean loosening governance. It means building an environment where plans test, learn, and adjust while keeping clinical oversight intact.
The plans that win here will be judged by fewer barriers and stronger relationships, not by tool count.
Looking ahead
The future of care coordination comes down to how well organizations combine technology, clinical expertise, and human judgment to build more connected care.
AI adds real capability, but its value shows up in how much more proactive, personalized, and responsive an organization becomes for the people it serves.
Keep the member at the center of every decision and care coordination stops being a back-office function. It becomes a strategic advantage, one that improves outcomes while reducing the total cost of care.
This article is inspired by the discussion and perspectives shared during the Care Coordination Reimagined for Cost Reduction roundtable at the Emids Healthcare Summit 2026.
Visit Emids Healthcare Summit page for more insights.
