<ul><li>This study investigates the protection offered by federated learning algorithms against eavesdropping adversaries.</li><li>The focus of the research is on safeguarding the client model itself.</li><li>The study examines various factors that impact the level of protection, such as client selection, local objective functions, global aggregation, and eavesdropper's capabilities.</li><li>The results highlight the limitations of methods based on differential privacy in this context.</li></ul>

On Model Protection in Federated Learning against Eavesdropping Attacks

Discover more