AI Adoption in Health Care Climbs, Led by Administrative Tasks

What Happened

Nearly 60 percent of medical organizations have adopted artificial intelligence, according to findings published in the 2025 CAQH Index, with administrative processes accounting for the largest share of that growth. The data, cited in a contributor report published by The American Journal of Managed Care, reflects a measurable shift in how health care organizations are deploying AI tools across their operations.

What the Data Shows

The 2025 CAQH Index tracks AI adoption across health care administrative and clinical functions. According to the report, adoption is concentrated in administrative workflows rather than clinical care. These include areas such as prior authorization, claims processing, and eligibility verification, which have historically been labor-intensive and paper-dependent processes.

The CAQH Index is an annual industry report produced by the Council for Affordable Quality Healthcare, a nonprofit alliance that works with health plans, providers, and other health care stakeholders. The Index has tracked administrative transaction volumes and cost data across the health care sector for multiple years, making the 2025 edition part of a longitudinal data series.

Why Administrative Processes Are Leading

Administrative tasks in health care represent a significant share of total operational cost. Prior research from CAQH has estimated that the health care industry spends tens of billions of dollars annually on administrative transactions, including manual phone calls, fax-based workflows, and staff time spent on eligibility checks and claim follow-up.

AI tools applied to these workflows can automate repetitive data entry, flag errors in claims before submission, and respond to routine payer inquiries without human intervention. Because these functions do not directly affect patient diagnosis or treatment decisions, they carry a different regulatory and liability profile than clinical AI tools, which may lower the barrier for organizational adoption.

Broader Context

The CAQH findings arrive as regulatory scrutiny of AI in clinical settings continues to develop. The U.S. Food and Drug Administration has received and reviewed submissions for AI-assisted diagnostic tools, including a generative AI radiology tool submitted by Aidoc that was reported separately. Clinical AI applications face a more demanding approval pathway than administrative software tools.

At the same time, payer and provider organizations have faced sustained pressure to reduce administrative overhead. The Centers for Medicare and Medicaid Services has introduced rules requiring faster prior authorization response times, a regulatory change that increases the operational value of automated processing tools.

Health care technology vendors including Epic, Oracle Health, and a range of smaller point-solution companies have introduced AI-assisted administrative features in recent years, contributing to the market conditions reflected in the CAQH data.

What It Means in Practice

For health systems and medical practices, the data suggests AI is being integrated into back-office operations at a pace that now involves a majority of organizations. The administrative focus means that many patient-facing workflows, including diagnosis support and treatment planning, remain largely outside the current wave of AI deployment.

For health plans, increased AI use by providers in claims and authorization workflows may alter processing volumes and error rates on the payer side, though the CAQH report does not quantify those downstream effects in the excerpted findings.

The contributor report in The American Journal of Managed Care notes that AI adoption is increasing broadly, but does not specify which organization types, such as hospitals, physician practices, or insurers, account for the largest share of the reported adoption figure.

The full 2025 CAQH Index report is expected to provide a complete breakdown of adoption rates by transaction type, organization size, and administrative function.