Two-thirds of health systems now deploy agentic AI to cut clinical burden

Adoption Milestone: Two-Thirds of Health Systems Turn to Agentic AI

Health-care providers face a mounting crisis: clinician shortages, administrative bloat, and fragmented patient data are stretching workflows to a breaking point. According to MIT Technology Review, more than 68% of health-care organizations have now turned to agentic AI—autonomous agents that can handle insurance processing, scheduling, and triage—as a strategy to collapse the cognitive load on medical staff.

The distinction matters. Prior waves of health-care digitalization—electronic health records in the early 2000s and remote monitoring tools—added friction rather than relief. Electronic health records remain fragmented and dependent on manual input. Telehealth improved geographic access but failed to match in-person care quality or build patient trust. Agentic AI operates differently: it makes autonomous decisions, queries clinical knowledge sources, and iterates without requiring human intervention for edge cases outside a rigid algorithm’s bounds.

Hospital for Special Surgery’s Insurance Claims Breakthrough

Ashis Barad, MD, chief digital and technology officer at Hospital for Special Surgery (HSS), an academic medical center in New York specializing in musculoskeletal conditions, framed the opportunity plainly: “Agentic AI takes your workflow and collapses it, augments it, supercharges it, and makes it more performant.”

HSS deployed AI agents to handle insurance claims—a process that previously spanned weeks and required both in-house staff and a third-party contractor to manage volume. The results, in nine months: AI agents now process 1,100 claims per month. Appeal processing dropped from 45 minutes to 5 minutes. Appeal success rates soared from 65% to 100%. HSS consolidated all claims work in-house.

Building on that success, HSS is now scaling to patient-facing workflows. In collaboration with enterprise agentic AI developer Ema Unlimited, the system offers 24/7 appointment scheduling and triage via web, text, or phone. The conversational agent asks patients clarifying questions about their condition, then matches them to the appropriate clinician while factoring in location, insurance coverage, and physician availability—completing, as Dr. Barad said, “the whole loop.”

Why This Matters

The 68% adoption threshold signals a turning point: agentic AI is no longer experimental. For health systems hemorrhaging administrative time, the economics are immediate—reduced appeal processing time and near-perfect success rates directly free clinicians to focus on patient care. The shift also reframes digitalization’s purpose: not to automate clinicians themselves, but to automate around them, removing procedural friction so that clinical judgment remains the bottleneck.

The question now is whether these gains hold as deployment scales beyond insurance and scheduling. If agentic AI can handle triage and front-office complexity reliably, the downstream implication is a rebalancing of clinician time toward diagnosis and shared decision-making—the work that patients value and that humans do best. Monitoring whether this actually occurs, and whether it closes gaps in underserved geographies, will determine whether agentic AI lives up to its promise of rehumanizing care.