The best AI will be the kind you don’t talk about: 5 healthcare AI predictions that will shape 2026 While many industries are experiencing AI fatigue and disappointment, healthcare remains a major exception. The need is real, the use cases are tangible, and momentum is going to keep building next year. That’s what makes looking ahead to the next year in healthcare AI genuinely exciting.I’ll admit last year’s predictions, I got a number of things right … and on a few fronts, I may have simply been a little early. That’s the nature of predicting technology curves in an industry that moves so fast. What’s different as we head into 2026 is the sense that experimentation is starting to give way to impact. I believe 2026 will be a year that challenges some of our assumptions about AI in healthcare – what it can do, what it can’t and where it delivers value. More importantly, I believe it will be a year of real progress. Here’s what I believe is on deck for 2026 (you can watch the video below, or keep scrolling to read through my five predictions): 1. It will be the ‘build-everything’ year (followed by a boomerang back) My first prediction for 2026 is that it’s going to be the year large healthcare enterprises truly believe they can build most, if not all, of their AI solutions in-house. As the technology improves and it becomes easier to spin up impressive demos, that belief will feel more attainable than ever. I actually see this as a bit of a pit stop – a year of experimentation that, in many cases, turns into wasted effort. It’s going to test vendors and startups alike, and it will require a real survival-of-the-fittest mindset to make it through. By 2027, I think we’ll see a clear boomerang effect back toward the vendor market, once organizations fully appreciate what it takes to build, scale, and maintain AI responsibly. 2. We’ll all get tired of ‘proactive’ “Proactive” will be among the most frequently used words describing AI in healthcare settings in the first half of 2026, and by the second half of the year, we’re all going to be completely exhausted by the word. Everyone is talking about how AI will transform reactive systems into proactive ones, and we’re going to hear that story over and over again. In fact, it won’t just become the most common word in AI storytelling – it might become the most hated one. The concept itself still matters, but the buzzword is going to lose its power as people demand more substance and fewer slogans. 3. It will be a year of dualities (delay and acceleration) I believe 2026 will be a year of real dualities, almost a tale of two AIs. On one hand, we should expect delays, particularly around data center construction, driven by a growing misalignment between supply and demand. Infrastructure simply hasn’t scaled as fast as the appetite for AI. On the other hand, adoption – especially in healthcare – is going to continue accelerating. 4. We’ll see the rise of invisible AI 2025 was the year AI agents dominated the zeitgeist. Every conversation, from the boardroom to individual contributors, centered on AI itself. In 2026, we’re going to pull back from that because people are burned out. Instead, the focus will return to what actually matters: outcomes, KPIs, and real-world impact. Whether something is powered by AI or not will become less relevant, and often not mentioned at all. The best AI will be the kind you don’t have to talk about. 5. True consumer engagement as a service finally arrives I believe we’re all about to finally see true consumer engagement as a service, particularly in healthcare. This is something the industry has talked about for more than a decade, but hasn’t been able to deliver well. AI is what changes that. With technology that finally has context and memory, we can begin to match the ambition of truly personalized healthcare engagement with real execution. What makes all of this especially exciting for me is that the Infinitus team will be right in the middle of these shifts. We’re working alongside healthcare organizations every day as they navigate what to build, what to buy, and how to deploy AI in ways that actually move the needle for patients and care teams. Seeing these trends up close gives me a lot of conviction about where the industry is headed, and how much real progress is possible. If you agree, disagree, or are seeing something different in your corner of healthcare, I’d love to compare notes. Feel free to reach out to me – these conversations are often where the most interesting insights emerge.