Website defacement attacks have been a significant threat to both private and public organizations’ websites and web portals. Such attacks can have severe repercussions for website owners, disrupting website operations and tarnishing their reputation, potentially resulting in substantial financial losses. With our approach, we examined SVM, which is a type of machine learning for detecting website defacement. Our approach applied machine learning methods to develop classifiers that distinguish web pages into normal and attacked classes. Moreover, we collected a large number of features from the website’s content and metadata to train and test the algorithms. This method is applicable to both static and dynamic websites; through training, it can learn to adjust to a wide range of page types. The use of an algorithm from machine learning to obtain results has shown that our approach achieves very high detection accuracy with a very low rate of false positives. Additionally, it should be noted that our approach does not require it to depend on massive computational capabilities.