Across healthcare, providers are rapidly deploying tools such as ambient clinical documentation, while payers are adopting AI-powered agents for a growing range of operational and administrative use cases.
Looking ahead to 2026, Becker’s Healthcare recently spoke with Jesse Cugliotta, Global Head of Healthcare & Life Sciences at Snowflake, about the current state of AI adoption in healthcare and how he expects trends like agentic AI, real-time AI capabilities and the use of data will evolve.
Note: Mr. Cugliotta will explore these predictions and related trends in greater depth during a webinar on January 15, 2026. More information here.
Responses have been edited for length and clarity.
Question: AI adoption in healthcare is occurring faster than previous technology waves. What’s driving this?
Jesse Cugliotta: Providers feel squeezed with patient volumes and acuity is higher, staffing shortages are continuing, there are huge administrative headaches and providers deal with reimbursement challenges. They are desperate and don’t have other levers to pull. So, they feel they have to move forward with AI. Yet even more important, AI solutions are providing real value for clinicians. That’s what’s driving adoption.
One of our major health system clients built an ambient listening technology on Snowflake’s AI framework. Physicians saw 15% more patients per hour and spent 2-3 fewer hours each day doing documentation. That’s the real value.
Q: We’re seeing growing interest in agentic AI. How might agentic systems support prior authorization workflows?
JC: So many discussions about AI in healthcare have focused on radical transformations. But the reality is agentic AI in prior auth, while not flashy, is much more achievable and able to deliver near-term value.
Prior authorization is a highly repetitive, rules-driven exercise that involves gathering data from multiple sources, validating it and providing supporting documentation. That’s exactly the kind of thing AI is really good at — it can understand rules, follow a series of steps and facilitate the workflow.
Snowflake has customers that have deployed AI to perform the steps associated with prior auth. Some customers have built agentic workflows where an agent auto-populates fields with information, auto-submits the data and handles some of the back-and-forth communication. We’re starting to see this accelerate.
Q: With the uncertainty surrounding ACA subsidies, how will payers and providers use data and AI to help navigate this uncertainty?
JC: If the subsidies expire, it will be extremely disruptive. The CBO predicted that about two million people who are mostly enrolled through ACA exchanges could drop their insurance. With similar disruptions in the past, it was not sick people who dropped their coverage; it was young, healthy people. When these individuals drop their coverage, it has an impact on risk pools.
When major disruptions occurred during Covid, payers took to the cloud to build analytics around real-time eligibility or coverage tracking to predict what their enrollments would look like and the impact on risk pools. Payers will need to do real-time risk scoring, which is an area where AI can play a big role. We also see AI-driven chatbots being used to answer members’ questions, decrease inbound calls and payers use it to personalize member experience.
For providers, this disruption could impact the services demanded. For example, elective surgeries may be delayed and EDs are likely to see higher volume and higher acuity. Providers are using data, analytics and AI to predict ED volume,patient flow and bed utilization and to forecast staffing — anything that can help optimize resources and improve productivity.
Q: As the demand for care and patient acuity rise, while staffing constraints persist, real-time insights for managing care and operations are increasingly important. What types of data capabilities will be most important to respond quickly and adaptively?
JC: Improving productivity is fundamentally a data problem. Healthcare organizations need to build an appropriate data foundation that leverages all relevant data to improve outcomes and productivity. This is not just data in the EHR; it is also clinicians’ notes, images, labs, -OMICS and more. It also includes data on social determinants and from outside of a health plan or health system. These require a strategy that makes it easy to integrate data from inside and outside of the organization without having to set up complicated pipelines.
This is an area Snowflake has always been good at. We arethe AI Data Cloud, which provides the ability to share data without having to copy it and physically move it.
Q: As AI capabilities advance, these tools and insights will become more widely available across organizations. How do you see this trend taking shape in the next year and what could it mean for how teams use information in their daily work?
JC: We believe AI allows healthcare organizations to leverage data at scale to produce insights and solve problems. We think of this as using AI and agentic tools to democratize critical health insights.
For example, one of our health system clients is using AI to improve infection control. Previously this health system had clinicians manually review charts and apply various rules to submit infection control information to the CDC and elsewhere. This process took 7-10 days.
Now, using a solution built on Snowflake’s Cortex AI framework, this health system has reduced clinicians’ chart review time by over 98% while achieving 99% accuracy, which is higher than the human process. This health system has said they can never go back to the old way of doing things.
This is just one example of a time-consuming manual process that is dramatically enhanced by AI.
Q: Are there any other trends on your radar for 2026?
JC: We are seeing tremendous demand from customers, whether they are payers, providers or life sciences organizations, that want more interoperability and more open data. These customers and technology vendors are raising interoperability with us because we are often the glue or the connective tissue that enables interoperability.
It is now much easier to interoperate among source systems than it used to be, and cloud data platforms are enabling that. The desire for interoperability and bringing data together is necessary for AI to use this data. This trend is gathering momentum and will be prominent in 2026.
