Cyber security is the information technology security that protects the system and interconnected users in the organization from several cyber threats and attacks. It aims to safeguard the sensitive and personal information of the users from external attacks and malware. The main significance of cyber security is to empower the confidentiality, integrity, and accessibility of data sources.

Compared to conventional learning concepts, Federated learning owns the high capability to be utilized in cyber security, beneficial to reduce cyber-attacks and to achieve data privacy and security. Federated learning is the decentralized training that collaboratively trains the model without exploiting the user data to preserve privacy. Federated learning for cyber security achieves data privacy, data fragmentation, data security, confidentiality, integrity, accessibility of data sources.

Applications of federated learning in cyber security are authentication, privacy preservation, trust management, intrusion detection, and anomaly detection. Application areas utilizing federated learning for cyber security are finance, the internet of things(IoT), edge computing, connected and automated vehicles. Federated learning for cyber security focuses on privacy preservation and indistinct to resist several kinds of vulnerable attacks, which become a challenge for federated learning. Currently, federated learning is improved with many techniques for providing solutions to attain privacy over different kinds of attacks in cyber security. Future scopes of federated learning for cyber security are efficient backward traceability, 5G networks, and hyperparameter tuning.