For years, the pitch for AI in healthcare has sounded almost too simple: sharper diagnoses, faster decisions, less paperwork. The real challenge isn’t picturing that future. It’s making it work inside a healthcare system that’s already dauntingly complex.
A patient does not experience healthcare as a single technology. Their information moves between providers, platforms and systems. A consultation can involve video, messaging, documentation and medical records, while behind the scenes, organizations have to keep that information secure and accessible.
That is starting to change the way the industry thinks about AI. The most consequential technology may not always be the application a patient or doctor sees. It can be the infrastructure underneath it: the systems that move information, support digital care and make increasingly automated workflows possible.
The investment numbers show a market still willing to bet on that future. Digital health startups pulled in $7.4 billion in venture funding during the first half of 2026, while McKinsey estimates generative AI could generate somewhere between $60 billion and $110 billion in annual value for healthcare.
The question now is what happens when that potential meets the reality of healthcare systems that were not designed around AI in the first place.
Building the infrastructure behind digital care
Telehealth offers a good example. The pandemic pushed virtual care into the mainstream, but the long-term challenge was never simply putting a doctor and patient on a video call. It was connecting that interaction to everything that comes before and after it.
A virtual consultation can involve patient intake, messaging, documentation, billing and medical records. Each step creates another point where information has to move between systems without becoming fragmented along the way.
That is the problem infrastructure companies are increasingly trying to solve.
QuickBlox, for example, provides communication infrastructure and a white-label telehealth platform that can be integrated into healthcare products rather than requiring providers to build an entirely new patient-facing system.
Its Q Consultation platform brings together secure video and messaging with digital intake, AI assisted workflows, documentation, and integrations with the healthcare systems already in place. The technology can also run through the cloud, a private cloud, or on-premises environments, depending on what a provider needs.
The distinction matters because healthcare has built up a complex technology layer over decades. A new AI tool might impress on its own, but its real value depends on whether it can work with the systems already handling patient information.
AI is making that integration problem more important, not less. As models move into intake, documentation and other routine processes, more healthcare workflows are beginning to depend on the infrastructure connecting them.
Data security becomes part of the AI equation
The more connected those systems become, the more the question of data matters. Healthcare organizations already manage huge amounts of sensitive information. Adding AI only adds another layer of complexity. Data now moves between applications, models and cloud environments. That means organizations still need to know where it lives, how it’s being used and who can access it.
The stakes are high. IBM’s 2026 Cost of a Data Breach research continues to place healthcare among the industries facing the highest costs when breaches occur.
That is creating demand for technologies focused not on the visible AI layer, but on the systems that allow healthcare organizations to understand and control their data.
Source Meridian is one example, focusing on healthcare data profiling and security. Its Healthcare Data Profiler recently achieved HITRUST e1 Certification, reflecting the growing emphasis on establishing clearer controls around sensitive healthcare information.
For the industry, this is becoming part of the same conversation as AI adoption. An organization can deploy an AI model. That alone doesn’t answer the harder questions about the data feeding it.
Those questions only grow more important as AI moves from isolated experiments into everyday workflows, touching patient records, communications and administrative processes.
The next healthtech opportunity may be underneath the application
Healthcare’s next technology cycle may therefore look different from the last one.
Instead of treating AI as another layer that can simply be added to an existing product, providers and technology companies are increasingly confronting the infrastructure required to make those systems work at scale.
That means better ways to connect patients and providers, move information between systems, understand where sensitive data is going and introduce automation without creating another set of disconnected tools.
The opportunity is less visible than the latest AI assistant, but potentially more consequential. Healthcare does not operate as a collection of isolated applications. It is a network of people, records, platforms and processes that have to work together.
As AI becomes part of that network, the infrastructure underneath it will determine how far the technology can actually go. The next stage of healthcare AI may be defined not simply by what the models can do, but by whether the systems around them are ready to support it.
