I had the opportunity to moderate a panel earlier this week on the top AI trends in healthcare, featuring speakers from the Mayo Clinic, Oracle, UCLA Health, and a former partner at Mass General Brigham’s AI and Digital Innovation Fund. It was part of the GAI World conference in Boston.
Here’s what was on their trend radar:
Tom Kingsley, Director of the Epidemiology AI Lab, UCLA Health
• Reducing wait times and increasing efficiency in large medical centers. “What we’re seeing is a lot of small hospitals closing, and access to care and wait times at larger centers like UCLA and Mayo Clinic and MGH are increasing,” he said. “Where I think we’re going to start to see in a lot of the investment in these tools [is to] address some of these major issues around access times and kind of the efficiency of care that we deliver as healthcare organizations.
• Maturation in the application layer. “How do we test these [AI applications] rapidly to figure out which ones are useful? How do we further develop them if they’re being internally developed?” Kingsley said.
Santiago Romero-Brufau, Director of Artificial Intelligence – ENT, Mayo Clinic
• “Agents are going to be the big topic,” Romero-Brufau said. “Very related to agents is this idea of context engineering, which is basically deciding what information you’re going to give the generative AI model, the LLM, in order to generate whatever output you want from it.”
• “I don’t think we’re going to see a lot of deployment of AIs for medical-decision making directly to consumers. I think that’s still a few years away. But [you will see it for] things like helping physicians with the notes, things like prior authorization, and kind of dealing with all that paperwork,” he said.
• Smaller healthcare institutions finally move to the cloud. “When I look at the numbers…a lot of the smaller institutions are not cloud-based yet. I think that is going to move very, very fast, as a lot of these [AI] models are essentially cloud-based. You can’t get access to GDP or Gemini or the equivalent, if you don’t have a strong cloud presence.”
Gaye Bok, Former Partner, Mass General Brigham AI and Digital Innovation Fund
• “The deployment of things that have been piloted…across health systems at scale is a trend I see in 2026,” especially in primary care and behavioral health, Bok said. “I don’t think it is coincidence that that is where there are very acute labor shortages in clinicians,” she said.
• “…Health systems are under a huge amount of financial stress. I think that is likely to intensify in 2026, so you’re going to see people really trying to look for how can they deploy these AI tools in the back office as much as possible.”
• “Typically, implementations and adoption [of AI in healthcare] have been slowed, because anytime that you ask for a change in workflow, that’s been a real barrier to adoption,” Bok said. “We might see that start to change where there’s more openness to [adapting] workflows to take advantage of these tools.” But Bok said that might not play out until 2027.
Christopher Boone, Group Vice President – Health and LIfe Sciences, Oracle
• Building agentic AI into the clinical workflows for healthcare professionals so that using it feels natural.
• Using AI and data to optimize the design of clinical trials.
• Dynamic value-based contracting, “which we’ve never really quite been able to do successfully before, but that’s one of the key things we’re working on,” Boone said.
You must be logged in to post a comment.