For the past two years, healthcare has been flooded with AI pilots that promise transformation. Many of them work, at least in controlled environments. But inside health systems, a different pattern is emerging.
Across health systems, leaders are discovering that the challenge is no longer identifying use cases or selecting the right models, it’s what happens after that decision is made.
AI initiatives that show early promise often stall when they encounter the operational complexity of real healthcare environments; fragmented systems, inconsistent data, and workflows that were never designed to support advanced analytics at scale.
This is not a new issue. In fact, it reflects a pattern the industry has seen before.
For years, healthcare has invested heavily in digital tools without fully addressing the foundational systems required to make those tools work together. AI is now exposing those gaps more clearly than ever. The more sophisticated the technology, the more dependent it becomes on the quality, accessibility, and interoperability of underlying data.
In other words, the most important work in AI isn’t happening at the surface. It’s happening underneath.
The ‘iceberg’ reality of AI in healthcare
There is a tendency to focus on the visible layer of AI, the algorithms, interfaces, and outputs that capture attention. But those elements represent only a small portion of what determines success.
Below the surface is everything that makes AI viable in practice: data standardization, interoperability, governance, security, and integration into clinical and operational workflows. When those elements are missing or underdeveloped, even the most advanced AI solutions struggle to deliver meaningful impact. Models trained on clean, curated datasets often encounter very different conditions when deployed in live environments. Inconsistent coding, incomplete records, and fragmented data sources can quickly degrade performance.
This is where many organizations hit a wall.
It’s not that the AI doesn’t work. It’s that the system around it isn’t ready.
From experimentation to execution
The industry is now entering a new phase, one where success will be defined not by innovation alone, but by execution.
Health systems are increasingly asking more pragmatic questions:
- Can this be integrated into existing workflows?
- Can it operate across multiple systems and settings?
- Can it scale beyond a single department or use case?
These are infrastructure questions, not algorithmic ones. And they require a shift in how organizations approach AI strategy. Rather than starting with the tool, leaders are beginning to recognize the need to start with the environment in which that tool must operate.
This includes investing in interoperability frameworks that allow data to move consistently across systems, establishing governance models that ensure data quality and trust, and building platforms that support continuous learning and improvement.
Without these elements, AI remains stuck in a cycle of promising pilots that never translate into enterprise value.
The persistent challenge of data quality
Amid rapid technological change, one principle has remained constant: garbage in, garbage out.
No matter how advanced an AI model becomes, its outputs are only as reliable as the data it receives. This is particularly challenging in healthcare, where data is generated across a wide range of settings, each with its own standards, systems, and levels of completeness.
Even regulatory approval does not solve this problem. A model that performs well in controlled conditions may behave very differently when exposed to the variability of real-world data.
Addressing this issue requires more than technical fixes. It requires a commitment to standardization and consistency at scale, something that has historically been difficult to achieve in a fragmented healthcare landscape.
Interoperability as a strategic lever
Too often, interoperability is framed as a compliance requirement or technical necessity. In reality, it is one of the most significant enablers of performance in modern healthcare systems.
Emerging research suggests that improving interoperability can have measurable impacts on system efficiency and financial performance. When data flows more seamlessly, organizations can reduce duplication, streamline workflows, and make more informed decisions in real time.
In the context of AI, interoperability becomes even more critical. It is what allows models to access the breadth and depth of data they need to generate meaningful insights, and to do so consistently across different care settings.
This is why interoperability should not be viewed as a prerequisite to innovation, but as a core component of it.
A path forward
The path to scalable AI in healthcare will not be defined by a single breakthrough or technology. It will be shaped by the industry’s ability to address the foundational challenges that have existed for decades.
This means shifting focus from isolated innovation to system-wide readiness. It means prioritizing the “unseen” work that enables everything else. And it means recognizing that the success of AI is ultimately tied to the strength of the infrastructure that supports it.
For healthcare leaders, this may not be the most exciting message. But it is an essential one. Because the future of AI in healthcare won’t be determined by what the technology can do. It will be determined by whether the system around it is ready.
Photo: Supatman, Getty Images
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
