Healthcare organizations have moved past the curiosity phase with artificial intelligence and are now focused on the harder work: actually scaling AI to deliver real financial and operational results. At this year's ViVE conference in Los Angeles, the shift was unmistakable. AI was everywhere, but the conversation had fundamentally changed. Rather than debating whether AI belongs in healthcare, executives are now wrestling with how to govern it, integrate it into daily workflows, and prove it generates meaningful return on investment (ROI). The challenge is significant. Many healthcare organizations have successfully piloted AI use cases, but translating those pilots into organization-wide programs that actually save money and improve care remains elusive. The gap between enthusiasm and execution is real, and it's forcing healthcare leaders to rethink their entire approach to AI adoption. Why Are Healthcare Leaders Shifting Their AI Strategy? For years, healthcare organizations treated AI as an innovation experiment. Today, it's viewed as essential to staying competitive and financially viable. However, executives have learned a hard lesson: AI won't deliver value if it's simply layered onto broken or fragmented workflows. Instead, value comes from redesigning operations and care processes so AI can be properly governed, scaled, and measured for real productivity, cost, and workforce impact. The focus has shifted dramatically toward practical, operational applications. More than 80% of health system executives are prioritizing agentic AI, which is a more advanced form of AI that can sequence tasks, independently respond to changing conditions, and orchestrate work across platforms to achieve clinical, administrative, and financial results. This is different from generative AI, which primarily responds to prompts. Similarly, 70% of health plans are prioritizing agentic AI for utilization management, prior authorization processes, and claims management. The early wins for AI are emerging in operations, where it can remove costs, improve workflows, and stabilize staffing. Healthcare organizations are using AI to improve revenue-cycle management, care coordination, workforce scheduling, and prior authorization processes. Some are even using it to enhance call center operations. What Are the Key Areas Where Healthcare AI Is Making an Impact? - Revenue-Cycle Management: AI is helping healthcare organizations streamline billing, reduce claim denials, and accelerate payment processing, which directly impacts cash flow and financial stability. - Documentation and Administrative Tasks: Technologies like ambient documentation can reduce the time clinical staff spend on administrative work by capturing insights from conversations that doctors might otherwise miss. - Workforce Scheduling and Staffing: AI is being used to optimize staff scheduling and help stabilize workforce challenges, which is particularly critical in healthcare where staffing shortages are endemic. - Care Coordination and Prior Authorization: AI can help coordinate care across departments and automate prior authorization processes, reducing delays in patient care and administrative burden on staff. One critical insight from healthcare leaders is that AI should be invisible to end users. While the technology has tremendous potential to reduce clinician workload, staff may resist it if they perceive it as an additional step in their workflow. The best AI implementations are those seamlessly integrated into existing systems like electronic health records (EHRs), dashboards, analytics platforms, and care workflows. How to Successfully Scale AI in Healthcare Operations - Redesign Workflows First: Before implementing AI, healthcare organizations must redesign their operations and care processes to ensure AI can be properly integrated, governed, and measured for impact rather than simply layering AI onto existing fragmented processes. - Integrate AI Into Daily Systems: Ensure AI is embedded directly into the tools clinicians and staff already use daily, such as EHRs and care coordination platforms, so it becomes part of the natural workflow rather than an additional burden. - Establish Clear Governance and Measurement: Define how AI will be governed across the organization and establish metrics to measure productivity gains, cost reductions, and workforce impact before scaling broadly. - Address Data Fragmentation: Connect fragmented diagnostic data across platforms so clinicians have a holistic view of patient information, reducing time spent reconstructing patient histories and enabling faster clinical decisions. - Build in Safety and Security: Ensure technology safety is built into AI systems from the start, especially for connected devices that generate or exchange clinical data, rather than treating security as an afterthought. Healthcare leaders are also recognizing that fragmented diagnostic data creates hidden costs. When imaging, pathology, oncology, and cardiology data exist on separate platforms, clinicians spend significant time trying to connect disparate data streams to reconstruct a patient's history. This operational burden delays clinical decisions and essentially functions as a tax on the healthcare system. "Value is unlikely to come from layering AI onto fragmented workflows or chasing shiny point solutions. Instead, value is likely to come from redesigning operations and care processes so AI can be governed, scaled, and measured for real productivity, cost, and workforce impact," noted Jay Bhatt, managing director of the Deloitte Center for Health Solutions. Jay Bhatt, Managing Director, Deloitte Center for Health Solutions What's Holding Healthcare Back From Realizing AI's Full Potential? Despite widespread enthusiasm, several obstacles remain. Questions persist about when AI will begin generating meaningful savings. The ROI could come from increased productivity, cost reduction, and workforce stabilization, but realizing that value requires more than just deploying technology. It requires fundamental rethinking of how healthcare organizations operate. Additionally, some AI solutions promoted at industry conferences don't actually address the challenges healthcare organizations are trying to solve. This mismatch between vendor offerings and real organizational needs means that healthcare leaders must be more discerning about which AI solutions to invest in and which to avoid. Another emerging concern is the role of consumer-focused devices like wearables and at-home diagnostic tests. While these devices show promise, data quality, governance, and interoperability remain significant prerequisites for their effectiveness. Healthcare organizations cannot simply adopt these technologies without addressing the underlying data infrastructure challenges. The path forward requires a long-term horizon. Lasting change in healthcare AI adoption involves demonstrating value over time and building trust among clinicians and staff. The critical question for healthcare executives is whether their strategies and approaches are driving toward scalability and sustainability, or whether they're chasing short-term wins that won't translate into lasting organizational benefit.
