GE HealthCare sits at the intersection of advanced medical technology, clinical workflows, and data-driven decision-making. The company operates across imaging, ultrasound, patient monitoring, and precision diagnostics, supporting healthcare providers in more than 160 countries and touching millions of patient interactions every year. As hospitals face rising cost pressures, workforce shortages, and growing data complexity, the role of AI is rapidly expanding from experimentation to operational infrastructure across the healthcare ecosystem.

Against this backdrop, GE HealthCare has been investing heavily in edge computing, cloud platforms, and multimodal AI to help clinicians make faster and more informed decisions at the point of care. From operating rooms and radiology suites to hospital operations and patient flow, the company is exploring how AI can reduce administrative burden, improve clinical workflows, and unlock new efficiencies for health systems.

In part 1 of this conversation, Parminder Bhatia, Chief AI Officer at GE HealthCare, and Cindi Howson, Chief Data Strategy Officer at ThoughtSpot, explored the shift from AI that assists to agentic AI systems that orchestrate work across complex healthcare environments while preserving trust, safety, and accountability. In this continuation, the discussion moves deeper into edge AI, workforce impact, and why 2026 may mark the moment AI becomes foundational infrastructure for healthcare organizations.

Edited Excerpts

Q: GE HealthCare spans multiple areas from radiology to ultrasound, and you recently showcased new concepts at CES. What did you preview, and what stood out most?

The key takeaway from CES is how we build systems that combine cloud and edge capabilities. We’ve been talking about Edge AI for some time, but what’s different now is the computing power available on edge devices. That allows us to deploy much more capable models directly at the edge, unlocking a new class of use cases.

Many healthcare scenarios require low latency, real-time feedback, resilience, and security. Edge AI is ideal for situations that need fast and reliable access to critical information, while still benefiting from the reasoning and interaction capabilities of agentic and large language models that pull data from multiple sources.

A helpful way to think about the balance between edge and cloud is air traffic control. An aircraft does not wait for a central system to respond to turbulence. It immediately takes local control in the cockpit, while broader coordination happens centrally. Healthcare needs a similar architecture.

One example we announced at CES involves anesthesia. The operating room is a highly constrained environment, and anesthesiologists manage rapidly changing physiology while monitoring vitals and administering anesthesia. They are sterile, time-constrained, and fully focused on the patient. Within that environment, imagine removing the legwork required to move across multiple screens and administrative processes so they can interact more directly with the patient.

Q: When you talk about Edge AI, do you expect smaller, more efficient language models to dominate in 2026?

It will be a hybrid approach. On one side, we are seeing significant advances in NPUs, GPUs, and CPUs at the edge. These devices can now host even larger models and deliver substantial reasoning capabilities locally.

At the same time, many use cases require multimodal interaction. Using the anesthesia example, a lot happens at the point of care during the operation. But there are also scenarios where you need to bring in multimodal data from EHRs, genomics and gene sequencing, and ER systems.

To generate better insights across these areas, you need the ability to pull large volumes of data and perform reasoning in the cloud. When we think about edge capabilities, the future is clearly a hybrid architecture where edge and cloud work together to solve complex problems.

Q: Let’s talk about job impact. Data, analytics, and AI leaders must help their teams embrace change across roles such as radiologists, anesthesiologists, and data analysts. What would you advise them?

AI will not replace jobs in healthcare or in general. However, it will reshape them positively.

Today, too much clinical time is spent on administrative friction rather than patient care. AI is exceptionally good at automating repetitive, high-volume, and bureaucratic tasks. By removing that burden, AI gives clinicians time back and creates the cognitive space needed to address burnout.

This shift allows physicians, nurses, technologists, and coordinators to operate at the top of their licenses, focusing on complex decision-making, clinical judgment, and meaningful patient interactions.

That makes one skill essential: domain understanding paired with critical thinking. AI can surface patterns and propose next steps, but humans remain responsible for applying context, synthesizing information, and deciding what to do next. AI helps propose, but humans decide.

Today, 80 to 90% of the time often goes into collecting and bringing data together from different places. Imagine using technology to unify those datasets. That unlocks new use cases that were not possible with siloed data in the past. It opens the door to solving higher-level decision-making problems using domain expertise.

To fully realize this shift, we must invest in people. That includes AI literacy, reskilling staff, and providing support as roles evolve. The future clinician is not replaced by AI; they are amplified by it through generative AI and advanced technologies.

Q: Looking ahead to 2026, what is the one thing leaders should pay attention to?

The biggest shift we will see in 2026 is moving AI from experiments to infrastructure. We need to start treating AI as an infrastructure investment. That is when real transformation begins.

AI is becoming embedded into the systems that keep hospitals running every day, and we are already seeing this play out in the real world.

For example, we worked with The Queen’s Health Systems in Hawaii to optimize patient flow using AI. In the first year, the initiative was associated with more than $20 million in savings by reducing the length of stay. That creates financial benefits for hospitals while improving care and access by freeing up beds for patients who need them.

At Children’s Mercy in Kansas City, similar approaches have been associated with more than an 80% reduction in patient admission delays and over an 85% reduction in canceled surgeries.

These improvements, shorter wait times, better care, and smoother workflows benefit clinicians, patients, and administrators. But capturing this value does not happen automatically. It requires clear ownership, strong operational discipline, and a mindset shift from asking whether we should use AI to focusing on how we solve the problem with AI.

Leaders need to set performance expectations, monitor real-world outcomes, and treat AI with the same rigor as any other critical hospital infrastructure. When AI is governed, measured, and embedded properly, it stops being a pilot and becomes something you can depend on. That is the difference between promise and impact, and we will see much more of that impact in 2026.

CDO Magazine appreciates Parminder Bhatia for sharing his insights with our global community.