Medicare AI Prior Authorization Pilot Criticized for Rushed Rollout and Patient Care Delays, New Documents Show

Newly released documents reveal that Medicare’s ambitious artificial intelligence pilot aimed at streamlining prior authorization processes was launched too quickly, resulting in significant delays in patient care. The program, overseen by the Centers for Medicare & Medicaid Services (CMS), sought to use AI tools from companies like IBM Watson Health and Cerner to automate approval decisions for medical treatments. However, internal reviews show that the system’s early deployment lacked sufficient testing, leading to inaccurate recommendations and bottlenecks that delayed critical care for beneficiaries—particularly in rural and underserved areas. Key stakeholders, including CMS officials and healthcare providers, have expressed concern over the lack of transparency and oversight during the rollout. The pilot, initially announced in late 2023 and implemented across select regions by early 2024, was intended to reduce administrative burdens on providers and improve access to timely treatment. Yet data from March 2024 indicates that nearly 15% of AI-assisted authorizations required manual override due to errors, with some patients waiting up to two weeks beyond recommended timelines. Providers reported frustration with inconsistent AI outputs and inadequate training materials, raising questions about the readiness of such systems for nationwide scaling. This episode underscores a growing tension between innovation and accountability in health tech. As federal agencies push forward with AI integration in public health programs, the stakes are high: missteps can compromise patient safety and erode trust. Moving forward, experts stress the need for rigorous validation, provider input, and clear regulatory frameworks before expanding automated decision-making in healthcare. The outcome of this pilot could shape future policy on AI in Medicare and beyond.