Healthcare organisations struggle to operationalise AI because leadership treats it as an IT problem
Healthcare organisations are moving beyond basic digitisation toward using AI for clinical workflows, analytics, and operational efficiency. Yet most struggle to scale these initiatives beyond pilot projects, according to Sathiyan Kutty, Chief AI Officer at Emids, a healthcare-focused technology and consulting company.
The barrier isn't technology. It's three things working against each other: unclear outcomes, fragmented data, and leadership that hasn't aligned around AI as a strategic priority rather than an IT purchase.
The three mistakes holding healthcare back
Kutty, who previously worked at Tesla and Kaiser Permanente, identifies a consistent pattern across organisations attempting to scale AI. First, companies fail to define what they're actually trying to accomplish. They hear about AI and try to apply it without contextualising it for their own business.
Second, data. AI requires large volumes of well-organised information. Many healthcare organisations either lack sufficient data or haven't structured it effectively.
Third-and largest-is organisational change. "Leadership teams tend to treat AI as an IT problem rather than a strategic priority," Kutty said. That disconnect between leadership, business objectives, and technology adoption is what holds most organisations back.
Healthcare AI adoption moves faster than other sectors
Healthcare is adopting generative AI and LLM technologies at least 2.2% faster than most other industries, according to Kutty. The shift accelerated after the arrival of large language models.
Before generative AI, AI for healthcare was primarily used in clinical diagnostics and research. Now it's embedded directly into workflows on the administrative side, where automation drives immediate efficiency gains. What would have taken years of traditional transformation is being accelerated by generative AI bridging existing technical debt.
The speed matters. What traditionally took six months of robotic process automation discovery now happens in twelve weeks.
Regulation demands responsible AI in healthcare more than other sectors
Healthcare in the US is actively adopting AI, but methodically. The industry is highly regulated, similar to financial services. Deterministic machine learning and deep learning will continue. Generative AI components are still in a test-and-go phase in many organisations.
"If there's one industry that truly needs responsible AI, it's healthcare," Kutty said. "Anything that touches human life demands the highest level of governance and guardrails."
Model drift-where AI system behaviour shifts when a new model version arrives-is a significant challenge. Many healthcare leaders don't yet know how to solve for it. Without proper guardrails, the consequences can be serious.
Leadership misconceptions about AI outcomes
The biggest misconception Kutty sees among leadership is that AI will solve problems without requiring internal change. Change management is consistently underestimated.
Some leaders demand outcomes without doing anything differently. Others treat AI as a forcing function-a multiplier for the entire organisation. They know how to build structures to support it.
The common blind spot at the early stage is thinking everything is possible without acting on it. "I can do it in ChatGPT, why can't we do it at scale?" the thinking goes. ChatGPT is a one-on-one relationship. When you introduce an organisational structure, cascading effects follow. Teams often lack project managers who know how to run an AI project.
The Chief AI Officer role is fundamentally a business role
Kutty defines success in the CAIO role as setting the right AI adoption strategy with clearly defined business outcomes, not pet projects. The role should not sit in IT.
A critical part of the job is identifying the pivot point where transformation becomes visible to the organisation. When the entire organisation is running at 20% to 50% more capacity and people actually feel the bandwidth difference, that's when the next level of adoption kicks in. Chief AI Officers have to know when their organisation is approaching that inflection point.
Near-term opportunities for payers and providers
For payers, prior authorisation and revenue cycle management have enormous potential. Longitudinal cost analytics-looking at member information across multiple providers and claims-was never practical without agentic solutions because the underlying data is so complex and unstructured.
For providers, AI adoption traditionally started in clinical settings. For payers, administration is the entry point. They're now connecting it back to predictive member analytics, identifying chronic conditions early and managing costs before they escalate.
Diabetes management is a concrete example. Intervene early, and the cost of managing that patient drops significantly. That benefits payers, members, and providers.
The future: AI becomes invisible, interoperability improves
AI is the buzzword today, but the future is ambient. Every product and experience will have an AI component woven into it. In a few years, people won't notice it any more than they notice electricity.
Kutty sees AI bringing healthcare organisations closer together than ever before. Contextual information sitting in payer, provider, and member systems couldn't talk to one another. Agents are beginning to change that. If one field is wrong in a data exchange, systems can now infer intention, parse it faster, and contextualise it faster. That's interoperability 2.0.
One issue that needs to change is data ownership. Patient data is currently held by large technology companies that want to remain monopolies. Patients should own their medical data and share it with whoever they choose for their own care. That's not the reality today, and it needs to be.
Your membership also unlocks:
