Present-era fake news is one of the major issues in society because it misleads everyone. Fake news involves a piece of a news topic that contains fake or misleading information. Similarity detection to a news writer’s point of view for a particular or against a particular topic is a complex task. For the dynamic evaluation of fake news detection, we proposed machine learning techniques, which will investigate the use of a combination of convolutional neural network (CNN) and long short-term memory (LSTM) models for identifying fake news and similarity detection. Using these models in the first stage the text data can be pre-processed and then sent to the next stage. The second stage extracts the important features from the text data using the CNN model. Finally, the sequential structure of the text data is formed with the help of the LSTM model. The output of these models shows certain binary classifications. Performance of evaluation metrics like accuracy, precision, recall, and F1 score demonstrates that the model is highly effective in detecting fake news using both CNN and LSTM.