Because fake news and clickbait can spread false information and cause havoc, they have grown to be major social issues. Machine-learning models have demonstrated great promise in identifying fake news, but adding clickbait detection to these models can improve their precision even more. In this study, we have investigated the use of various machine learning models to identify clickbait and identify fake news. First, we have evaluated how well the well-known machine learning algorithms Naive Bayes, SVM, and Long Short-Term Memory (LSTM) performed at identifying fake news. To train and test these models, we used a dataset of news articles both with and without clickbait. With an accuracy of 88%, our results demonstrated that LSTM performed the best when using clickbait. The best performing model, LSTM, was then combined with clickbait detection to create a hybrid model. We extracted clickbait elements from the headlines using regular expressions and combined them with the content of the articles. Our findings demonstrated that the LSTM model had the highest accuracy out of all the models.