According to a detailed review of government documents obtained through litigation and reported by the Electronic Frontier Foundation (EFF), Medicare’s Wasteful and Inappropriate Service Reduction (WISeR) program, which uses AI to process prior authorization requests, has encountered significant issues including delays, denials, and patient harm within its initial months of operation.
- Widespread delays and improper denials reported in early months
- Financial incentives may encourage vendors to reject prior authorization requests
- Transparency remains limited on AI systems and their accuracy safeguards
Best for / avoid if
The WISeR AI prior authorization model may be considered relevant for health systems or policymakers seeking to explore AI-enabled cost control mechanisms in Medicare or large-scale insurance programs. Its ability to quickly process requests has theoretical appeal for reducing administrative burdens when functioning correctly. However, the current evidence suggests that healthcare providers and patients affected by this system should be cautious about relying on WISeR, given documented delays, frequent erroneous denials, and adverse effects on timely care.
Organizations prioritizing patient safety, equitable access, and transparency should avoid adopting similar AI authorization systems without robust safeguards, clear auditability, and strong human oversight. The program’s early experience illustrates risks when financial incentives align poorly with care needs and when AI decision-making lacks public scrutiny. For patients requiring urgent or complex treatments, this technology in its current form may pose unacceptable delays and barriers.
Pricing and alternatives to check
CMS’s WISeR program compensates contracted vendors primarily based on denied prior authorization requests, with penalty reductions for poor quality scores capped between 5-10%. These modest financial penalties do not currently appear sufficient to dissuade denial-heavy practices or ensure optimal patient outcomes. Precise pricing tiers or alternative commercial plans are not detailed in the released documents, limiting direct cost comparison analysis for buyers or stakeholders.
Potential alternatives to WISeR’s approach include more transparent prior authorization models that emphasize timely human review, enhanced data sharing, and independent AI audit frameworks. Other AI-assisted prior authorization programs outside of Medicare or private insurers focusing on collaborative clinician involvement and strict bias mitigation may offer different cost-benefit balances. Buyers are advised to scrutinize transparency, oversight capacity, and incentives when considering AI authorization tools.