Internal corrosion in pipelines is governed by a multitude of factors, such as the chemical composition of fluids, microbiological activity, operational conditions, material factors and the elevation profile. Complex relations between those factors make it difficult to develop a comprehensive theoretical model. Additionally, operators often do not have sufficient data collected for such a model. This study introduces a novel phenomenological model which utilizes inline inspection (ILI) data and other factors, if available, to identify regions of pipelines that have a high probability of a release due to bottom-side internal corrosion. Utilizing data from clients in addition to public release data from PHMSA, we developed and validated a model for pipelines with confirmed releases due to internal corrosion. A release score was developed, based on the fraction of pinhole-sized anomalies, local and global depth outliers and factors such as anomaly orientation, presence of active corrosion, and the depth of local elevation minima. The model successfully predicts release sites in several pipelines with known releases. This predictive capability supports targeted maintenance and preventive measures for improving pipeline integrity management. Future work will focus on refining the model based on updated release data and availability of measurements for other factors mentioned above.