Healthcare organizations are pushing AI adoption faster than their infrastructure can support, according to Nutanix's latest Enterprise Cloud Index healthcare report. The findings reveal 88% of healthcare IT leaders say their current systems are not ready for on-premises AI workloads, while 79% of organizations report employees deploying unsanctioned AI tools across clinical and administrative functions - a combination that puts patient data, security, and clinical outcomes at risk.
The report, based on a global survey of IT and engineering leaders, shows AI deployment in healthcare is being driven from the top even as frontline systems remain underprepared. Up to 75% of healthcare data is expected to be generated at the point of care, intensifying the need for low-latency, secure infrastructure that can support real-time AI applications in clinical environments.
Infrastructure gaps at the bedside
High-density care settings such as intensive care units can involve up to 20 connected devices per bed. Each device generates data that AI systems could use for early warning scores, treatment recommendations, or resource allocation. But when infrastructure cannot process that data locally, delays and system failures become clinical liabilities, not just IT headaches.
The shift from centralized data processing to bedside applications changes the stakes entirely. Daryush Ashjari, chief technology officer and VP of Solution Engineering, APJ at Nutanix, said the tension is palpable. "Healthcare organisations across APJ are under growing pressure to adopt AI, but clinician demand is colliding with the readiness of the infrastructure underneath it. The priority is to build a unified, hybrid approach that bridges data sovereignty requirements with the need for real-time insights at the patient's bedside."
Shadow AI spreads through clinical and administrative teams
While infrastructure lags, staff are not waiting. The report finds that 79% of organizations are encountering shadow AI - employees deploying unsanctioned tools without IT oversight. A majority of respondents, 83%, view these activities as a business risk. Persistent silos between IT and operational teams make coordinated responses difficult, leaving organizations exposed on multiple fronts.
For healthcare leaders working through governance and strategic infrastructure decisions, the gap between executive AI mandates and frontline realities continues to widen. AI for Executives & Strategy resources can support leaders navigating these cross-functional challenges, where clinical urgency, compliance requirements, and technical constraints must be balanced simultaneously.
Containerization emerges as a bridge
The report points to a structural shift in how healthcare organizations modernize their application environments. Some 86% of respondents cited containerization as a key enabler for deploying secure, portable workloads closer to where data is generated. This approach supports compliance with data sovereignty requirements, which 72% of organizations now consider essential.
Adoption is expected to scale quickly. More than half of respondents anticipate running at least five AI-enabled applications within three years, spanning generative AI, predictive analytics, and autonomous agents. The path from ambition to clinical impact runs through infrastructure modernization - not around it.
As healthcare organizations scale their AI deployments, the pressure to train clinical and operational staff on both the tools and their governance is intensifying. Targeted AI for Healthcare programs address the practical knowledge gaps that shadow AI exploits, from data security protocols to evaluating AI outputs in patient-facing contexts.
Why this matters for healthcare professionals
Shadow AI does not stay in the back office. When clinicians use unsanctioned tools for documentation, triage suggestions, or treatment research, the output enters the patient record - and the liability chain - without validation, audit trails, or security review. The infrastructure investment conversation is not just a capital planning issue. It determines whether the AI tools that reach the bedside are governed or ad hoc, integrated or bolted on, auditable or invisible. Healthcare professionals should push for clarity on which AI tools are approved, how they were evaluated, and what infrastructure sits behind them before relying on their output in clinical decisions.
Your membership also unlocks:
