Healthcare leaders may still be looking for ROI and sustainability, but they know that AI will stay here and they want to make sure they are on the right path.
Healthcare executives say the hype around AI is justified, but amid all the epic declarations and catchy metaphors, they are still trying to figure out where technology fits into clinical workflows.
“We’ve seen the hype, but we haven’t seen much material,” said Nasim Eftekhari, executive director of Aiphied AI & Data Science at City of Hope, who attended the recent Vive 2025 event and participated in the AI Innovation panel.
Certainly, one of the bigger takeaways from Vive was that AI is still all the rage, but in all pilot programs and early use cases, it is still difficult to find value in clinical care. While the REV cycle and the financial sector are aiming to win by reducing administrative tasks and calculating numbers, clinical leaders are trying to figure out how to integrate AI into the care pathway.
And for all health systems and hospitals, someone else has seen their value as they have achieved success with ambient AI for doctors and messaging in baskets. If you’ve seen one successful use case, there’s an old proverb, but you’ve seen one successful use case.
But with HIMSS 2025 opening this week, healthcare leaders want to talk about AI again. And they largely agree that the technology will affect the industry in a serious and substantive way.
“There’s a lot of AI in everything,” said Simon Nazarian, EVP of Hope’s EVP and Chief Officer of Digital and Technology, who spoke at Vive. “But a lot of that is good old automation.”
So, where is the next step? And do we really need an a-ha moment to move things forward?
Technology and clinical care have a complex history, as veterans and “meaningful use” prove to be evidenced. By drawing from these memories, executives are tentative to fully embracing AI and really want to see what it can do before making a commitment.
That’s nothing new and it won’t scare executives from using AI. The agenda for this week’s HIMSS 25 meeting and Sunny (hopefully) the Las Vegas exhibition in Las Vegas is filled with discussions about examples of health systems for testing AI and technology.
For many, catchwords are now governance. It is understood that AI, in particular generative and predictive models, evolve as more data is collected. This means healthcare leaders should be aware of what comes up on the other side, not just how they collect data and feed it to the machine.
In the case of MS’s Dr. John Haramka, president of Mayo Clinic Platform, AI “will become everything we do.” Speaking on Vive, he said Mayo is testing some generative AI models. All of these are designed to enhance clinical care rather than replace clinicians.
Halamka said that healthcare systems and hospitals can launch all the AI tools they need, but they can better understand the outcome.
“As long as you understand the risk of breaking things, you can move quickly and break things,” he pointed out.
That being said, Halamka knows that AI is transformative. The best endoscope on the planet still misses 15% of the small polyps, he pointed out, but the AI tools developed at Mayo missed 3%. And he added that AI is part of the standard of care and hospitals can sometimes be sued for medical malpractice because they don’t use it.
It means it takes time to get there from here. There is a mistake.
Eftekhari said in City of Hope that AI will save lives. Predictive tools can help clinicians find health concerns faster and identify the best route of care.
“Responsible use of AI has the ability to move upstream,” Nazarian added, noting the possibility that AI in the study could ultimately identify and further prevent cancer. “We want to go from the bench to the bed as soon as possible.”
So, in all of the conversations at HIMSS this week, in conversations about how AI is being tested or used, executives need to understand that whatever they see or hear doesn’t necessarily work for them. The story of one meeting from one meeting to the next is probably the same. They need to shape the technology into specific goals.
Halamka says the industry needs to see both victory and failure, and they need to learn from each. He expects progress to be slow and stable, and hopes that the hype will continue until the use case is proven.
“Next year, we’ll talk about the truth,” he said.
Eric Wickland is Associate Content Manager and Senior Editor for Health Leader Innovation.
