The expansion of modern retail destinations in Saudi Arabia has increased the complexity of tenant and lease administration, particularly where rental arrangements include turnover-based components linked to tenant sales. Conventional tenant sales auditing depends heavily on periodic declarations, manual reconciliation, sampling, and retrospective verification. These practices can be difficult to scale across multi-mall portfolios and may delay the identification of abnormal sales patterns. This study proposes an intelligent tenant sales auditing framework that combines machine learning-based anomaly detection, retail sales forecasting, tenant benchmarking, contractual data analysis, and risk-based audit prioritisation. The proposed framework integrates tenant point-of-sale information, declared sales, historical performance, footfall indicators, lease conditions, payment records, and tenant-category benchmarks. A simulated evaluation demonstrates how the framework can classify tenants according to anomaly risk and support targeted audit intervention. The study does not claim that simulated results represent actual Saudi mall performance; they illustrate the analytical logic and potential operational value of the proposed approach. The framework contributes to research at the intersection of artificial intelligence, retail analytics, and commercial real estate management by shifting tenant sales auditing from periodic and reactive verification toward continuous, data-driven monitoring. The proposed model can support Saudi retail property operators in improving sales transparency, strengthening lease governance, prioritising audit resources, and developing more intelligent turnover-based lease management practices.