September 2026 | By Lindsey Miller
3 Takeaways for Health Plans:
- Generative AI is changing health plan solution design by giving business and clinical leaders a more direct role in defining and developing technology. Beyond reducing administrative burden, the opportunity is to use the speed to improve how care is administered and delivered.
- AI can reduce administrative burden and accelerate health plan operations, creating more capacity for work that requires human expertise and judgment.
- Connected, agentic AI workflows offer greater value than isolated tools by carrying context and work across systems, teams, and the member journey.
If you have worked in technology long enough, you have probably seen the familiar tree-swing illustration. A business leader asks for a swing, the request moves through layers of interpretation and development, and the final product looks very different from what was originally needed.
Most of us have experienced some version of that disconnect.
The distance between the people experiencing a problem and the people building the solution has long-shaped enterprise technology. Business leaders explain what they need, requirements are developed, and technology teams translate those requirements into solutions.
Months later, the business may find itself adapting its processes and workflows around what was built.
Generative AI is changing that dynamic. Business leaders can describe a problem conversationally, pressure-test a workflow, refine business rules, explore exceptions, and help develop the foundation for a prototype or technical requirements. Technologists remain critical to the process, while the people closest to the work gain a much more active role in defining what gets built.
That shift can give health plans new speed. Its value, however, should be measured by more than how much administrative work it removes. The larger opportunity is to use that speed to improve how care is administered and delivered.
What Does AI Mean for Health Plans?
At the ACHP Symposium in Washington, DC, I had the opportunity to speak with health plan and industry leaders about how their organizations are using AI. One of the first things that became clear was that we were not always talking about the same thing.
AI has become an umbrella term. One organization may be talking about machine learning or predictive models that have been operating for years. Another may mean employees using generative AI tools such as ChatGPT or Claude. Others are beginning to explore agentic workflows capable of completing a series of tasks and involving a person when judgment or an exception requires attention.
Those distinctions have practical implications. A tool that summarizes a document carries a different level of responsibility than an agent capable of initiating a transaction, changing a member record, or moving work from one system to another.
Health plans are also having these conversations as AI moves further into implementation. McKinsey reported in April 2026 that 50% of surveyed U.S. healthcare organizations had implemented generative AI, up from 25% in its fourth-quarter 2023 survey. Agentic AI remains earlier in its development, with 19% reporting implementation and another 51% pursuing proofs of concept. The survey included leaders from payers, clinical-care organizations, and healthcare technology companies.
As adoption grows, health plans need greater precision around the problems AI is being asked to solve. That means understanding the decisions involved, the information required, the actions the technology can take, and where human judgment remains essential.
How Generative AI Is Shortening Healthcare Development Cycles
One example shared at the symposium made this shift tangible.
Dr. Amy Collins, Senior Vice President and Chief Medical Officer at CommunityCare in Oklahoma, shared how she approached an operational challenge involving prior authorization. Medical directors had been pulling cases from a queue, and Collins wanted to create a push model that could distribute work based on case complexity, physician availability, and practical considerations such as a medical director returning after a partial day of paid time off.
Those details may seem small, but they reflect the operating knowledge that can easily get lost as a technology request passes between teams.
Collins used ChatGPT to describe the desired workflow, incorporate lessons from a similar tool used at a previous organization, and develop technical requirements. Her technology team then used Claude to support development. The result, she said, was a working tool in about two weeks. A similar effort at her previous organization had taken nine months and still did not fully address the original business need.
The speed is impressive, but I think the most important change is who was able to participate in designing the solution.
“It really puts the subject matter expert just very much closer to the solution to the problem.”
Dr. Amy Collins, Senior Vice President and Chief Medical Officer at CommunityCare
That is the shift I find most interesting; not simply that AI helps produce requirements or code. But a clinical business leader translated firsthand operating knowledge into a solution in just weeks.
Why AI Could Give Community Health Plans an Advantage
Community-based health plans have historically faced a technology disadvantage when compared with larger organizations. They may have fewer specialized resources and less capacity for large enterprise implementation programs.
But they have another advantage: placing the subject matter experts much closer to the solutions.
Clinical and operational leaders are closer to the people performing a process every day. They understand why work gets stuck, where an exception creates three additional steps, and which seemingly minor detail determines whether a workflow succeeds.
Generative AI can make that knowledge more actionable.
A community plan with strong domain expertise may be able to define, test, and implement a focused solution faster than an organization carrying a larger technology stack and more layers of decision-making.
Scale will continue to matter. So will strong technology partners, infrastructure, governance, and technical expertise. But size alone does not determine which organization understands an operating problem best.
AI creates an opportunity to turn that understanding into a solution faster.
Can AI Reduce Administrative Burden for Health Plans?
Speed also raises an important question: What does a health plan do with the capacity AI creates?
One health plan leader discussed using technology to make medical directors more efficient. The goal was not to reduce the number of clinicians. The plan saw an opportunity to redirect their time toward work where their expertise could have greater impact.
That could mean spending more time on complex clinical decisions, improving consistency across reviewers, educating nurses, engaging providers, or providing more personal support. This capacity can be used to improve decisions and give clinicians more time for work that requires judgement, experiences, and empathy.
Administrative efficiency is already a major area of interest for healthcare AI. In McKinsey’s research, healthcare leaders identified administrative efficiency more frequently than any other area as having significant potential for both generative AI and multiagent workflows.
Reducing administrative work can produce meaningful savings, particularly for health plans operating under continued cost pressure. The larger opportunity is deciding where that newly available capacity can create the most value.
If technology gives a clinician another hour in the day, the measure of success should include what that clinician can now do with it.
Why Connected AI Workflows Matter for Health Plans
Softheon has been thinking about automation and agentic processes long before the current generative AI cycle.
Health plan operations depend on work moving between systems, teams, and people. Eligibility, enrollment, billing, payments, reconciliation, member communications, and service operations rarely exist in isolation. A change in one part of the member journey often creates work somewhere else.
Generative AI can add intelligence within the workflows, but its value increases when the relevant context and next steps can move with the work. It can help business leaders articulate requirements, give technology teams a stronger starting point for development, interpret information, and increasingly help determine what needs to happen next.
Members do not experience a health plan’s technology stack; they experience whether their coverage is active, their payment was processed, or their authorization was resolved in time. When systems and teams do not connect, staff chase information across tools while members repeat their story or wait for someone to connect the dots. The larger opportunity is to remove those seams before back-office friction interrupts coverage or care.
How Should Health Plans Measure the Value of AI?
The tree-swing problem has always been a problem of distance.
There is distance between the person experiencing the problem and the person defining the requirements. Distance between those requirements and the technology being built. And the distance between an operational improvement and the member it is ultimately intended to serve.
AI gives us an opportunity to shorten that distance.
Business and clinical leaders can participate more directly in solution design. Community-based plans can turn operating knowledge into focused solutions more quickly. Technology teams can spend less time interpreting what the business meant and more time engineering solutions that work within the larger operation.
Development time and administrative savings are useful measures, but they do not tell the entire story. We should also ask whether members receive answers sooner, clinicians have more time for difficult decisions, providers encounter less friction, and problems are resolved before they interrupt care. AI’s larger opportunity is to give the people who understand healthcare’s problems a more direct role in solving them and a clearer path to improve the experience of the people those systems ultimately serve.
I’m continuing the conversation around AI, health plan operations, and what it means for the future of healthcare. Connect with me on LinkedIn to follow along and share what your organization is seeing.
Ready to explore what AI could look like in your operations? Submit a request to connect with the Softheon team to learn how Softheon Habitat connects point AI solutions across workflows, systems, and teams.
