Case in point: professional healthcare equipment provider Phillips. Philips writes that healthcare AI is moving into a more active phase, with agentic AI beginning to support clinicians across everyday workflows. In the article, Philips describes AI agents as systems that can work inside existing clinical tools, understand clinical context and help coordinate tasks across teams while leaving medical decisions to healthcare professionals.

The clearest example is radiology, where AI agents can help with work before and after image interpretation, such as preparing patient summaries, organizing information and surfacing missing details. Philips frames this as a way to give radiologists more time for judgment, communication and patient care as imaging demand grows.

The larger point from Philips is that agentic AI could become part of a “hybrid healthcare workforce” that helps hospitals manage rising demand, workforce pressure and more complex care. Philips points to early adoption by health systems such as Mount Sinai and Mayo Clinic, as well as a National Health Service initiative in the U.K. focused on responsible deployment.

The article also connects agentic AI to two broader healthcare trends: using long-term patient data to spot risks earlier and bringing AI directly into procedure rooms to help clinicians see, guide and coordinate complex interventions in real time. Taken together, Philips’ view is that agentic AI will be most useful when it reduces busywork, connects information and helps clinicians act sooner without replacing their role in diagnosis or treatment.

Hitting the Pain Point Trifecta

GE HealthCare is also bullish on agentic AI. It argues that agentic AI could help solve three of healthcare’s biggest operating problems: too much data, too little coordination and systems that do not work well together. A recent article on the company’s website says clinicians are being asked to make complex decisions with information spread across lab results, images, patient histories and medical records.

GE HealthCare points to cancer care as a clear example. An oncologist may have only 15 to 30 minutes with a patient but still needs to review test results, imaging, medications, prior procedures and other health conditions before making a treatment decision. Agentic AI could help by pulling that information together, flagging what is urgent and giving care teams a clearer view of the patient.

The most important agentic AI idea in the GE HealthCare article is the “virtual tumor board.” In this model, different AI agents handle specific tasks, such as reviewing clinical notes, reading imaging data, checking biopsy reports, analyzing genomic results and helping schedule follow-up care. A coordinating agent then brings those findings together for the oncologist to review. GE HealthCare is careful to stress that oversight remains essential.

These systems would need strong privacy, security, audit and clinical review controls, especially because AI can produce inaccurate information. Still, the article’s larger point is optimistic: agentic AI could turn fragmented healthcare workflows into more connected care journeys, helping clinicians move faster, reduce manual work and give patients more timely treatment.

What Happens After Automation?

A recent feature in the Economist focuses on data and efficiency. The article frames agentic AI as a way to give time back to healthcare workers by taking on the manual tasks that slow care. The article points to appointment scheduling, medical data transfers, insurance coding, documentation and summarization as everyday sources of friction. Those tasks do not just add cost. They also contribute to burnout and can pull clinicians away from patients. KPMG’s Beccy Fenton says the biggest benefit is time, with one KPMG analysis suggesting that AI could cut time spent on some administrative tasks in half and return about 30 minutes per day to each doctor.

The article’s agentic AI focus is on what happens after basic automation. Older systems follow preset rules. Agentic AI can work toward a goal, such as scheduling a CT scan at the nearest in-network provider or deciding when a doctor needs to review medication changes.

The article says nearly 70% of healthcare organizations already use agentic AI in some form, although most efforts remain limited to specific workflows. The next stage will require cleaner data, stronger governance and systems that can safely connect across older technology. The message is practical: agentic AI could reduce administrative overload and improve patient care, but only if healthcare organizations build the data, security and oversight needed to let agents act reliably.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.