The Role of Federated Learning in Improving Financial Security: A Survey

TL;DR

Federated Learning offers a privacy-preserving path for finance to combine insights across institutions without sharing raw data. The paper introduces a regulatory-exposure taxonomy (Low/Moderate/High) for FL applications and surveys practical deployments in credit risk, portfolio optimization, loan predictions, AML, and fraud detection, including blockchain-integrated FL and FML-XAI. It highlights that FL can match centralized performance in some AML and credit-risk cases while enabling cross-institution collaboration, yet faces data heterogeneity, adversarial threats, and regulatory challenges. Looking ahead, the authors advocate for blockchain integration, differential privacy, secure multi-party computation, and quantum-secure FL to enhance privacy, auditability, and scalability in financial systems.

Abstract

With the growth of digital financial systems, robust security and privacy have become a concern for financial institutions. Even though traditional machine learning models have shown to be effective in fraud detections, they often compromise user data by requiring centralized access to sensitive information. In IoT-enabled financial endpoints such as ATMs and POS Systems that regularly produce sensitive data that is sent over the network. Federated Learning (FL) offers a privacy-preserving, decentralized model training across institutions without sharing raw data. FL enables cross-silo collaboration among banks while also using cross-device learning on IoT endpoints. This survey explores the role of FL in enhancing financial security and introduces a novel classification of its applications based on regulatory and compliance exposure levels ranging from low-exposure tasks such as collaborative portfolio optimization to high-exposure tasks like real-time fraud detection. Unlike prior surveys, this work reviews FL's practical use within financial systems, discussing its regulatory compliance and recent successes in fraud prevention and blockchain-integrated frameworks. However, FL deployment in finance is not without challenges. Data heterogeneity, adversarial attacks, and regulatory compliance make implementation far from easy. This survey reviews current defense mechanisms and discusses future directions, including blockchain integration, differential privacy, secure multi-party computation, and quantum-secure frameworks. Ultimately, this work aims to be a resource for researchers exploring FL's potential to advance secure, privacy-compliant financial systems.

Figures (2)

• Figure 1: A visual representation of the federated learning protocol b3. Each round of federated learning starts with a global model with initialized weights. Then, the FL server sends the global model to the clients. Next, the clients train locally on silo or device specific data and submit local updates only to the FL server. Then, the FL server performs some aggregation method on those local models to update the global model. Finally, the FL server sends the aggregated model back to the clients.
• Figure 2: A visual representation of the money laundering schemes. Image credit: Linkurious, available at https://linkurious.com/money-laundering-schemes/