Halfway through 2026, health tech executives predict what’s to come

Healthcare Brew has been covering developments in the health tech space closely this year, reporting on everything from policy changes to the rise of AI to big tech’s partnerships with hospitals—to name a few.

As we reach the halfway mark of 2026, we checked in with health tech experts about how they think the year’s been going so far, and what they’re anticipating over the next six months.

These responses have been lightly edited for length and clarity.

Taha Kass-Hout, global chief science and technology officer, GE HealthCare

What are you seeing as the biggest trends of this year?

The biggest trend is that AI is moving from pilots and point tools into real clinical and operational workflows. For the last few years, most organizations have been experimenting with AI. Now the conversation is much more practical. Healthcare systems are asking: How do I reduce burden on staff, improve throughput, shorten time-to-report, manage capacity, and help clinicians make better decisions?

The AI projects getting traction are the ones embedded into the workflow. In imaging, that can mean helping acquire a scan faster, improving image quality, automating measurements, prioritizing cases, or assisting with reporting. In operations, it can mean helping hospitals manage patient flow, staffing, scheduling, asset utilization, and capacity constraints.

What are you expecting for the end of the year?

I expect to see continued movement from pilots into production environments.

Healthcare organizations are becoming more selective. They are looking for AI solutions that demonstrate measurable value, fit naturally into workflows, and can be deployed at scale.

I also expect to see broader adoption of generative AI for assistive tasks such as documentation, reporting support, information retrieval, and workflow automation including staying up to date with guidelines, literature and evidence.

Does that align with what you thought would happen?

Yes, the industry is moving beyond proving that AI works technically. The focus is now on how AI performs in real-world environments, how it is monitored over time, how updates are governed, and how accountability is maintained when AI becomes part of clinical decision-making.

The future is beyond building more models. It is about building trusted systems with appropriate governance, monitoring, evidence generation, and sustainable economics.

Jahangir Mohammed, founder and CEO, Twin Health

What are you seeing as the biggest trends of this year?

We are seeing a shift from episodic care to continuous understanding. The healthcare system was built around isolated touchpoints: an appointment, a lab result, a prescription refill. For cardiometabolic disease, this approach leaves hundreds of days between visits where disease progresses unchecked. What’s changing is the recognition that continuous insights and personalized support can drive better outcomes.

At the same time, employers and health plans are demanding stronger evidence that programs are making a real difference. They’re asking whether A1C is actually coming down, whether members are coming off medications, and whether those numbers hold up in independent claims data. The bar has moved from participation to reversal.

The second trend we’re seeing is the emergence of GLP-1 stewardship. As demand for GLP-1 medications continues to accelerate, organizations are asking more rigorous questions about the appropriate duration of therapy and what a responsible exit looks like. Organizations are starting to take a responsible, outcomes-focused approach, evaluating who is most likely to benefit, how treatment should be managed, and when it may be appropriate to reduce or discontinue therapies.

Prashant Samant, co-founder and CEO, Akido

What are you expecting for the end of the year?

By the end of the year, I expect the conversation to become much more practical. The question will shift from “which model is best?” to “which organizations can safely use AI to expand care capacity?”

That means more focus on deployment architecture: how the AI is supervised, how clinicians stay in the loop, how liability works, how quality is measured, how reimbursement works, and whether the system produces better access and outcomes.

I also think human-in-the-loop deployment will increasingly be seen as a feature, not a limitation. In medicine, the goal should not be to remove clinicians from the system. The goal is to let clinicians safely supervise much more care, with AI doing more of the cognitive and operational work around them, so they can focus on what matters most: their patient.

Does that align with what you thought would happen?

Broadly, yes. It was predictable that the largest frontier models would become extremely strong at medical reasoning. Medicine is knowledge-heavy, language-heavy, and pattern-heavy, so it’s a natural domain for these systems to perform well in.

What we’ve always known is that “good at medical reasoning” is necessary, but not sufficient. Clinical care is asking the right questions, noticing what’s missing, building the right differential, deciding what needs to happen next, getting the patient to follow through, documenting correctly, billing correctly, and making sure someone owns the decision.