Start Small to Scale AI in Managed Care: Leanna Baker, PharmD, MS

Managed care organizations hoping to gain meaningful value from artificial intelligence (AI) should start with simple applications that establish the governance and structure needed to tackle bigger challenges later, Leanna Baker, PharmD, MS, digital solutions lead at Cencora, said in an on-site video interview at the Academy of Managed Care Pharmacy Nexus 2026 meeting in San Diego, California.

The interview followed Baker's session, "Is Your Organization Ready for AI? A Managed Care Playbook," on October 27, which presented primary research on AI adoption trends, perceived value, barriers, and gaps across managed care. The research covered ethics and governance, prior authorization, drug utilization review, and administrative burden reduction.

Baker recommended that managed care organizations begin with what she described as “low-hanging fruit,” such as automating review summaries or retrieving information from internal sources. Although these applications may seem basic, they help organizations build the governance and capacity for larger, more complex tasks as their programs mature, Baker explained.

The session framed AI adoption as a progression through 3 phases: piloting, scaling, and optimization. Baker said she does not believe any organization has reached optimization yet but noted that Cencora is in the scaling phase, applying value frameworks, decision processes, and model cards to the tools it has invested in to define how each will be used. This structure allows tools to progress from pilot to minimum viable product to finalized product, with ongoing maintenance and monitoring. In contrast, most organizations are still in the piloting phase, determining how to realize value from their existing investments, Baker said.

Central to that process is the value framework outlined in the session. Before developing a tool, organizations should define the questions they want to answer and establish a framework that includes key performance indicators, goals, expected efficiency gains, and clear criteria for success, she explained. Once a tool is in use, teams can continually revisit the framework to assess whether it is meeting its targets.

“We’ve already put together the metrics that we want to hit before we’ve even started, and we’re coming back to that value framework to say, ‘Are we hitting those metrics? Is this a successful pilot? Should we continue to do this? Is the investment too great, or should we move on to something else?’” Baker concluded.