The objective of the study is to develop and validate a model for classifying the technical condition of gas pipeline sections using in-line inspection (ILI) data and the Random Forest algorithm. Based on a dataset of 15,361 ILI records, a seven-parameter Random Forest model was developed in the GasPipelineInsight software; the data were stratified into an 80/20 train-test split and additionally evaluated using five-fold cross-validation. The classification accuracy on the test set was 98.31%, with a macro-F1 score of 0.92; the most significant contributions were made by allowable pressure, defect length and depth, the safe-service-life coefficient, and failure pressure. Data from Supervisory Control and Data Acquisition (SCADA) systems and the Internet of Things (IoT) were not used for model training or testing and are considered a prospective temporal layer for subsequent updating of the technical-condition forecast. The practical significance of the study lies in ranking pipeline sections according to potential risk and supporting decisions on inspection and repair.