Medicare WISeR AI Pilot Triggers Preapproval Errors Across Six States

The pilot applies AI-driven prior authorization to 13 services previously exempt in traditional Medicare. Oklahoma patient Bill Curry required three trips instead of one for an epidural, while contractors in Arizona, New Jersey, Ohio, Texas, and Washington reported incomplete systems still adding features months after go-live. KFF Health News documented confusion over documentation rules and extended wait times for approvals that Medicare had avoided for decades.

Rapid rollout from June 2025 announcement to January implementation bypassed standard testing periods used in prior CMS models. State medical associations noted providers received minimal training, echoing earlier private-insurer automation failures where false denials reached 20-30 percent before manual overrides. The program targets fraud in high-growth areas such as skin substitutes, whose spending rose nearly 700 percent, yet applies the same logic to routine pain procedures without published validation data.

Systemic risk emerges when unproven algorithms govern safety-net populations already facing access barriers. Historical patterns from Medicaid managed-care expansions show similar tech deployments increased administrative burden on rural clinics by 15-25 hours weekly. Without published false-positive rates or appeal outcomes, WISeR risks shifting costs onto vulnerable patients and small practices rather than curbing misuse.

Next steps require CMS to release quarterly denial and appeal metrics by September 2026 and commission an independent RCT comparing automated versus manual review before national expansion.