Natural language processing (NLP) has impacted the area of information security in different ways. NLP is widely used to detect anomalies and fraud in financial documents. Different NLP techniques such as named entity recognition (NER), sentiment analysis, text classification, word embedding, keyword extraction and similarity detection are applied to detect financial document fraud. Further large language models (LLM have also opened a different era of security mechanisms by providing comprehensive and reasoning-based outputs. This chapter aims to highlight different anomaly and fraud detection techniques using NLP. Other security measures which can be taken by using NLP are also discussed. Further integration of machine learning and deep learning-based algorithms with NLP is also described to enhance security in financial documents of large structured and unstructured data sets. Different supervised, unsupervised, and semi-supervised machine learning models are being applied to detect anomalies and fraud in financial documents by learning from past data. NLP-based implementation using machine learning is also elaborated. Case studies are presented to gain insights into NLP in detecting financial fraud. Finally, lessons learnt are summarized with a conclusion.