During the period of digital communication, Global societies face serious issues as a result of the rise of fake news. Proliferation of misinformation impacts public opinion and disrupts democratic processes, emphasizing the need for robust detection methods. This study identifies and categorizes news stories as either authentic or fraudulent using several machine learning (ML) and deep learning (DL) techniques. Several algorithms were evaluated depending on how well they identified bogus news by use of a large dataset of labelled news instances. Logistic Regression achieved an accuracy of 94.47%, while Multinomial Naive Bayes attained accuracy of 84.35%. Support Vector Machine (SVM) classifier gained an accuracy of 96.13%, and Random Forest classifier attained an accuracy of 92.07%. Decision Tree attained an accuracy of 87.96%, and XGBoost achieved an accuracy of 95.79%. Convolutional Neural Networks (CNN) performed well with an accuracy of 99.50%. Long Short-Term Memory (LSTM) gained an accuracy of 91.34%, and the hybrid approach combining CNN and LSTM executed with the highest accuracy of 99.59%. Among the tested algorithms, the hybrid method combining CNN and LSTM attained the maximum accuracy, closely followed by CNN. These outcomes demonstrate the promise of the tested algorithms in providing accurate and reliable fake news detection, offering valuable insights for the development of automated tools to combat misinformation.