FOIA records reveal delays and testing gaps in Medicare's AI prior auth pilot
Newly released Freedom of Information Act records expose significant operational failures in the Centers for Medicare & Medicaid Services (CMS) Wasteful and Inappropriate Service Reduction (WISeR) pilot. The roughly 1,000-page document set, obtained by the Electronic Frontier Foundation, details how the AI-assisted prior authorization program, launched in six states this January, suffered from…
Key points
- FOIA records show Medicare's WISeR AI pilot had severe delays, with one prior authorization request unanswered for 83 days.
- Vendor Innovaccer warned CMS its system was untested and lacked full functionality weeks before the January launch.
- Vendors issued over 20,000 non-affirmations in the first three months, with Virtix denying more requests than it affirmed.
Internal communications reveal that vendor Innovaccer warned CMS weeks before launch that its system lacked full functionality and had not undergone sufficient end-to-end testing. Despite these warnings, CMS proceeded with the rollout. In the first three months, vendors Virtix Health and Zyter issued over 20,000 non-affirmations, with Virtix denying more requests than it affirmed. While some providers have since received "gold card" status to bypass the process, the records highlight a lack of transparency regarding AI training data, bias testing, and appeal outcomes.
The pilot is scheduled to run through 2031, but the findings raise concerns about potential expansion to urgent care services like MRI scans and oncology. Critics argue that the current system's instability poses risks to patient safety, particularly if applied to time-sensitive medical procedures without first resolving the existing delays and technical gaps.
New Records Reveal Backlogs, Testing Gaps in Medicare's AI Prior Authorization Pilot
medscape.com · 13 September 2026
Loading the full article…
This text was published by medscape.com and written by Steph Weber. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Policy & Regulation
All →- Anthropic gives six METR evaluators employee-like access to internal systems · 8 src
- Netflix documentary reveals OpenAI AI agents breached Hugging Face in July 2026 · 4 src
- White House AI task force starts 120-day review of superintelligence risks · 1 src
- Anthropic adds US voting data from Democracy Works to Claude before midterms · 3 src
- FTC subpoenas OpenAI and Anthropic over AI safety incidents · 6 src
Comments
via GitHub Discussions