Despite widespread adoption of artificial intelligence tools across the healthcare payer landscape, the vast majority of organizations are failing to realize meaningful returns on their investments. According to a new study by Everest Group supported by Sagility, 84% of payer executives report limited benefits from AI implementations, even as 70% of U.S. health plans have already deployed AI tools for administrative and clinical activities.

The disconnect between adoption and impact reveals a fundamental challenge in how healthcare organizations are approaching automation: they're layering sophisticated technology onto fragmented legacy systems rather than re-engineering the underlying workflows first.

The Automation Paradox

The research paints a sobering picture of AI's current state in healthcare operations. Only 10% of health plans have observed measurable operational improvements from their AI investments, suggesting that most organizations are experiencing what the study characterizes as "automated chaos"—where AI flags unnecessary claims reviews because the underlying contract terms or prior authorization rules remain flawed.

The problem stems from operational silos that characterize most payer organizations. Claims handling, utilization management, payment integrity, and member services typically operate on separate platforms, preventing AI from delivering the integrated intelligence needed to drive real efficiency gains.

Anand Biradar, Chief Go-to-Market Strategy and Enterprise Transformation Officer at Sagility, emphasizes that the solution requires integrating clinical context, contract logic, and claims processing into a unified ecosystem. This approach enables AI to function across the entire claims lifecycle—following a Prevent-Detect-Resolve/Recover-Learn feedback loop that addresses root causes rather than symptoms.

The Human-AI Partnership Model

Notably, 94% of payer executives emphasize the need for human oversight in automated tools, pointing toward an "augmented intelligence" model rather than full automation. This framework positions AI as an intelligence amplifier for registered nurses, certified coders, and claims specialists, enabling these professionals to focus on complex clinical decision-making while AI handles data synthesis from unstructured medical records.

The stakes are considerable. The U.S. healthcare system spends approximately $1 trillion annually on administrative costs driven by manual processes, complex clinical reviews, and lack of systems integration. Without workflow re-engineering, this massive spending category remains largely unoptimized despite growing AI investments.

Implications for Provider Operations

For healthcare providers, the implications extend beyond payer operations. Workflow integration promises to reduce provider friction by minimizing unnecessary claim denials and streamlining prior authorization processes. When AI operates within properly engineered workflows, it can reduce first-submission errors and decrease the administrative burden that contributes to provider burnout.

The research suggests that healthcare organizations need to resist the temptation to pursue quick AI wins through point solutions. Instead, the path to meaningful ROI requires the harder work of redesigning fragmented workflows, integrating disparate systems, and then deploying AI within that optimized infrastructure—a sequence that reverses how most organizations have approached their digital transformation efforts to date.

Sources

  1. Re-Engineering Workflows to Bridge Healthcare's AI ROI Gap — HIT Consultant