Intermountain Health is leaning heavily into automation and data-driven tools to offset clinician shortages and simplify high-volume work across the enterprise with about 300 AI projects underway, according to Dan Liljenquist, Intermountain’s chief strategy officer.
The accelerated focus on AI will ensure the system is ready to overcome industry headwinds and expand access to care.
“Menlo Ventures just published a report a couple of weeks ago about the adoption of AI in different industries, and healthcare is one of the fastest adopters of AI,” he said during an interview with the “Becker’s Healthcare Podcast.” “It couldn’t be more timely. The whole industry is in flux and this massive demographic shift that we’re seeing as a society at the same time of clinician burnout is at all time highs. AI actually gives us some pretty interesting new tools to help us simplify really the work that needs to occur for our patients.”
Intermountain’s team has taken a pragmatic approach to innovation and adoption, understanding the urgency of transformation. Mr. Liljenquist said vendor partners are accelerating, too.
“We’re seeing some pretty fast adoption, partly because we’re building right off of the platforms that we’re currently using,” he said. “It’s kind of exciting for us, and this is one of those areas where there’s actually a tailwind for the industry instead of a headwind.”
The efficiency gains are already visible. Intermountain has deployed an AI agent to streamline administrative work across the health system, including the revenue cycle department. Managing claims denials and appeals was once a labor-intensive task that took a great deal of time; that’s no longer the case.
“Drafting appeals letters is a great example. We’re taking about thirty minutes off each appeal letter we have for a payer because of our ability to use AI to scrape through and get the information we need for the appeal letter,” said Mr. Liljenquist. “That’s just one of dozens and dozens of different uses that are just making our jobs simpler and easier to do.”
Enabling a tech-first workforce transformation
With the high volume of AI-driven projects Intermountain is doing, the system is experiencing a paradigm shift from a labor-first to tech-first revenue cycle. This transition is critical for all health systems and payers to lower the $1 trillion-plus spent on healthcare back office and administrative work.
“The revenue cycle is a repetitive task and rules based. AI will do a great job sitting into that space,” he said. “But we also think that there’s opportunities in analytics and call centers and supply chain. Think of your repetitive tasks that are rules based. Those are the opportunities where AI can help us move forward.”
Intermountain’s cloud migration also opened the door to faster development cycles and dramatically lower cost structures.
“We’ve moved all of our data into the cloud, and because we’re in a cloud environment, it’s a secure environment. We’re on Microsoft Azure,” said Mr. Liljenquist. “We’re able to actually use these tools, a lot of the OpenAI code, etcetera, to go directly into our data and build new applications using Vibe coding and other kind of pretty fascinating techniques and capabilities that AI can bring to do things that that would take a programmer weeks of time. It takes the ROI to that type of initiative down to near zero.”
Incumbents have a data advantage — but there are risks
Large and established health systems have a meaningful advantage in the AI race because they already hold the structured and unstructured data required to train safe, accurate models. They’ve built trusting relationships with legacy vendors to stay ahead of the innovation curve. But that doesn’t mean new entrants and disruptors are completely locked out.
“Incumbents will have an advantage for a while, partly because the data already exists inside the data to train AI is already inside our firewalls, inside our environments,” he said. “I think we have an advantage for a time if we lean in. Short of that, the real risk to the industry is a complete disintermediation between the system-patient relationship with a new startup that does something better, more effective, and gains the trust of the patient. We don’t have the luxury of sitting back and just waiting for things to happen. I think we’re going to have to lead it.”
It’s not enough to just collect the data and organize it neatly. Health systems need experts to look into processes and uncover data sources that will help identify potential biases and issues ahead of creating AI models.
“To make AI safe for use and to really make it effective, you’ve got to actually understand the underlying data used to create the outputs you’re looking for,” he said. “If you train AI on bespoke datasets, you get bespoke AI instances that may not lead you into the right future. Really understanding your data and having that data organized, the barrier to entry for a lot of these AI applications are really, really small. But the one barrier you have to overcome is data, and the incumbent systems have the data.”
AI governance is now a board-level conversation
AI growth has required new structures for oversight and risk management. C-suites and boards are creating committees to develop governance and assess risk with the right experts and stakeholders from departments systemwide. But, the top leaders also need a deep understanding of AI governance and risk to make the right strategic decisions.
“We have dozens of different projects coming each month through a governance process and, depending on the level of risk, that escalates up all the way up into our board conversations, board level approvals for certain uses of AI,” Mr. Liljenquist said. “But not every use is really that controversial. What we’ve gotten really good at, and we’ll continue to get better, is streamlining and understanding the use case we’re trying to build and scoping the risk based on the use case.”
The pressure to scale AI is tied directly to the workforce outlook. Intermountain’s C-suite is spending time identifying how AI can support faster and lower cost work. Their focus on the workforce is intentional and urgent.
“We’re an industry that’s largely built on labor,” he said. “That labor market is getting tighter and tighter. Over the next five years, many of the people we’ve built this model around are retiring, and there’s not enough people coming back through to replace them. We’re anxious to push the boundaries of what AI can do so we don’t have to shrink back from meeting the needs of our communities. Instead, we’re looking to expand what we can do.”
AI will reshape the practice of medicine
Beyond the administrative burden, AI will change how patients receive care and how clinicians practice, especially as shortages worsen.
“I think AI is going to change the practice of medicine significantly,” Mr. Liljenquist said. “The biggest thing you’ll see is, over the next five years, a quarter of our providers in the United States are gonna enter retirement by 2040 or 2035. Forty percent of those providers are gone, right, at a time when the demand for healthcare services is skyrocketing.”
Medication titration as one example where AI could strengthen outcomes, reduce variability and expand access.
“The idea that a doctor is the only person who could do medication titration for your blood pressure medication or that you need to see a doctor every year to renew your prescription. Those models are breaking,” he said. “People will not have access like they have. I think we are excited at Intermountain to lean into those types of interventions.”
Intermountain is exploring new ways to use AI to improve situational awareness and support ongoing care.
“We’re really excited about leaning in to find new ways to have AI help us be much more situationally aware of what’s happening with our patients, help them get to the right stable medication doses, and not require them to come back and see a doctor every time they have a tweak in their medication,” he said. “We see AI as a way to extend what we’re doing in very low cost ways to better meet the needs of our community, to be proactive for them, to simplify their experience, and help us partner with them better across the course of their lives.”
