With a rapid increase in new technology, follows new cybersecurity challenges. Federated Learning is a new type of Machine Learning, first proposed in 2016. End clients work together with or without a central server to train a Machine Learning model. A Federated Learning model trains sub-models on each client and later sends the model parameters back and forth, either between each other or the central server. This enables companies and organizations to work together without sending any of their private data. Much research has gone into detecting cyberthreats in this new type of Machine Learning; however, there is a lack of research on how to prevent cyberthreats from affecting the global model. This paper proposes and investigates a new approach for detecting and preventing several different types of poisoning attacks from affecting a centralized Federated Learning model via average accuracy deviation detection (AADD). By comparing each client’s accuracy to all clients’ average accuracy, AADD detect clients with an accuracy deviation. The implementation is further able to blacklist clients that are considered poisoned, securing the global model from being affected by the poisoned nodes. The proposed implementation shows promising results at detecting poisoned clients and preventing the global model’s accuracy from deteriorating.