Abstract
Considering the extensive utilization of social media platforms, false information is more likely to spread rapidly among users. With the surge of user-generated content, a significant amount of the information being shared is in-accurate and has no connection to reality. It can be chal-lenging to automatically identify and verify the accuracy of textual content. This research project aims to employ B-LSTM deep learning model with dynamic and static pre-trained embeddings to classify news datasets utilizing three fake news benchmarking datasets.