Financial technology enables institutions to serve customers around the world 24/7. Its services are often readily available and allow customers to transact in real time. Due to these advantages, financial technologies are becoming increasingly popular among clients. As financial technologies transactions consist of information, ensuring security becomes a critical issue. Vulnerabilities in such systems expose them to fraudulent acts that cause serious harm to both clients and providers. For this reason, Machine Learning techniques are applied to identify anomalies in financial technology applications. They focus on suspicious activity in financial data sets and generate models to predict future fraud. We contribute to this important issue and provide an evaluation of anomaly detection methods in this matter. Experiments were performed on several fraudulent datasets from real and synthetic databases. The obtained results confirm that ML methods contribute to the detection of fraud with varying degrees of success.