AI in Action: What Hospital Leaders Are Learning

June 18, 2026

At HAP’s 2026 Leadership Summit, a panel of health care leaders explored how artificial intelligence (AI) is transforming patient care and hospital operations, sharing lessons learned from implementing the technology across their organizations.

Moderating the discussion, Jason Mitchell, MD, executive vice president and chief medical officer at Geisinger and a HAP board member, emphasized that AI is no longer a future concept.

“AI is already here and we are using it,” Mitchell said. “The conversation is how do we use it effectively to advance the care of the people we serve?”

Throughout the discussion, panelists agreed that successful AI adoption depends on thoughtful implementation, staff engagement, and a clear focus on improving patient care. Here are some key takeaways.

Ensuring thoughtful implementation

For Heidi Beilis, MD, FACR, vice president and chief medical officer of the diagnostic service line at WellSpan Health, success with AI has depended on careful testing and thoughtful implementation.

“We piloted everything before we implemented it,” Beilis said.

WellSpan’s earliest AI applications focused on breast cancer prevention and diagnostics. As the technology expanded, integrating AI into clinical workflows became a priority. Beilis said the impact was immediate, improving both productivity and workplace culture.

Changing behavior and workflows

Cara Martino, DNP, RN, senior vice president of clinical improvement and transformation at Jefferson Health, discussed the experience of implementing ambient AI technology. Jefferson uses AI to summarize clinical notes. Today, the system works very well but successful adoption required adjustments to workflow and behavior.

“We don’t nurse out loud,” Martino said, explaining that clinicians had to adjust their workflows to ensure the technology captured the information needed to generate accurate documentation. “That change in practice to start saying it out loud has been the biggest barrier. But when we do it, the technology works really well.”

‘It takes trust’

Richard Medford, MD, FAMIA, FIDSA, FRCP(C), chief digital information officer at Allegheny Health Network, said AI-powered scribes have become an effective tool when there is support from clinical teams.

“It takes trust,” he said. “The technology is there and works well.”

C-suite too must be comfortable taking financial risks with AI innovation, he said. Out of ten AI experiments, eight might lose money but two may produce significant operational improvements and savings that can be built upon.

“There has to be that pool for ‘failing fast,’ that pool for innovation,” Medford said. “Otherwise you’re doomed to failure.”

Creating buy-in from the front lines

Mark Sendak, MD, MPP, population health and data science lead at the Duke Institute for Health Innovation and co-founder of the Health AI Partnership, said skepticism among staff can be one of the biggest early hurdles for hospitals implementing new AI.

“There was a lot of concern about how AI was being pushed to the front lines,” Sendak said, adding that engaging front-line teams in designing AI implementation generates buy-in. “As an innovator, I never want to be an evangelist to the front-line staff about why they should use this.”

Hospital leaders addressed those concerns through transparency and communication, helping staff understand how the technology could support their work and improve outcomes.

“It has to be clearly understood how it solves problems,” he said.

Sendak encouraged organizations to identify best practices, pilot new technologies, and carefully evaluate how successful programs can be scaled across their systems.

See more key takeaways from Day 1 and Day 2 of HAP’s Leadership Summit.