Healthcare Tech in 2026: The Hard Part Isn't the Technology

Hospital vital signs monitor displaying heart rate, blood pressure, SpO2 and other patient data waveforms

A few things happening in healthcare technology right now that are worth paying attention to, wherever you sit.

A breach that should worry everyone

This week we learned the Oracle Health breach - the one involving legacy Cerner servers - touched close to 20 million people, based on what the company told the Texas attorney general. A federal judge already ruled a lawsuit against Oracle Health and eight health systems can move forward.

Think about that. Hospitals did not choose to get breached. Their vendor did. But the hospital is the one patients are angry at, and the one sitting in court.

That changes the math on security staffing for every health system watching. It is not enough to have a security team that watches your own network. You need people who understand vendor risk, legacy system exposure, and clinical downtime - because when a hospital's systems go dark, ambulances get diverted. That is not an IT metric. That is patient care.

The government is forcing systems to talk to each other

A federal rule finalized in 2024 (CMS-0057-F, if you like acronyms) is getting real teeth in 2026. Payers now have to open up APIs for prior authorization, and by January 2027 they have to support payer-to-payer data exchange.

This is unglamorous work. Nobody writes breathless headlines about FHIR endpoints. But it is real, it has deadlines, and it needs people who have actually built integrations in healthcare before. Demand for HL7 and FHIR experience is not hype - it is compliance with a clock on it.

AI scribes: real, but keep your head

The most genuinely useful AI in healthcare right now is ambient documentation - software that listens to the visit and drafts the note. The big systems are past piloting. Cleveland Clinic published its enterprise rollout playbook. Houston Methodist is deploying across the whole organization. One large study after a year and 2.5 million uses found doctors genuinely saved time.

But read the fine print. An independent study this spring called the time savings "modest." This is not magic. It is a tool that works when you deploy it carefully, train people on it, and check the output.

That is the pattern with clinical AI in general right now. HIMSS surveyed health systems this year and found a lot of organizations stuck - interested in AI at scale, but without the governance or the people to get there. The technology is not the bottleneck. The execution is.

A strange moment for health IT people

Here is the part worth watching closest if you work in or hire for this field.

Health system IT departments are going through a shake-up this year. Some are outsourcing and cutting. At the same time, hiring leaders say the hardest roles to fill are the hybrid ones - someone who knows AI AND security, or Epic AND clinical workflows. One health IT leader put it plainly: the skill they cannot find is not AI or security alone. It is someone fluent in both.

So you get layoffs in one building and unfilled critical roles in the building next door. The market is not weak. It is mismatched.

What this means in the Triangle

Raleigh-Durham sits in the middle of all of this. Duke, UNC and WakeMed run some of the biggest health technology teams in the Southeast. The Park is full of companies building the software those systems buy.

For employers, the practical read is that hybrid skills take longer to find than a job posting suggests - searches that start early close better. For healthcare IT people, the market signal is that cross-domain experience (Epic plus security, data plus clinical, integration plus compliance) is where the leverage is right now.

Neither of those is likely to change before the year does.