The increasing frequency and severity of network intrusions globally poses a significant threat. Unauthorized access to valuable network resources, often referred to as network intrusion, has the potential to greatly compromise security and data protection when unauthorized penetration into a digital network takes place. To address these attacks, an automated classifier is used to identify network intrusions. The main objective of the proposed solution is to develop an intrusion detection system that utilizes a combination of a Gradient Boosting classifier and Principal Component Analysis (PCA) for the purpose of feature extraction. By employing Principal Component Analysis (PCA), it is possible to convert a dataset with a high number of dimensions into a lower-dimensional representation, thereby reducing the number of features. This approach enables the system to effectively detect U2R and R2L attacks, as well as normal instances. The model was tested using the NSL-KDD dataset and achieved an accuracy of 99.61%. Additionally, the model’s performance was assessed by measuring its F1 Score, precision, and detection rate. The experimental outcomes obtained from utilizing the NSL-KDD dataset indicate that the model performs better than previous models developed for U2R and R2L attack detection. The proposed method offers a fast and precise solution for detecting network intrusions, presenting a novel approach to maintaining network security. Implementing such measures is essential in protecting valuable data from the growing prevalence and advanced nature of cyber threats.