This study explores prescriptive appropriateness in the Marche Region of Italy. Using NotebookLM, an AI-based methodology was developed to analyze 36,465 dematerialized prescriptions from 2025 across three critical clinical areas (ophthalmology, breast diagnostics, and electromyography) against RAO priority criteria. The results revealed that 30% of these prescriptions were inconsistent with RAO guidelines, with 55% of all anomalies concentrated among only 15% of the prescribers. Rather than a systemic failure, this digital audit identifies specific prescribing patterns, promoting clinical and professional accountability. The successful local pilot led to the regional platform established by DGR 885/2026. It aligns with the national PNGLA 2025-2027, which mandates diagnostic details and leverages AI tools to monitor clinical demand, ensuring healthcare equity, safety, and long-term sustainability.