Federated learning and Quantum Computing for cybersecurity in the banking Sector: a systematic review

Abstract

The modern banking sector faces growing
cybersecurity challenges, demanding collaborative yet privacypreserving
defense mechanisms. Federated Learning (FL) has
emerged as a key paradigm for addressing these needs, but the
advent of quantum computing threatens its long-term viability.
This development introduces a critical duality: the existential
threat of quantum capabilities to break classical cryptography and
the transformative opportunities offered by Quantum Machine
Learning (QML). While research has addressed these aspects
separately, a unified analysis that unites them is lacking. This
paper presents the first comprehensive systematic review mapping
this dualistic landscape. Following PRISMA guidelines, the review
synthesizes the current state of research and evaluates the
challenges and opportunities at the intersection of FL, quantum
computing, and financial cybersecurity. The findings reveal
a strong consensus on the urgent need to integrate Post-Quantum
Cryptography (PQC) to secure FL frameworks against future
threats. Simultaneously, a long-term vision is emerging to leverage
QML to enhance threat detection capabilities beyond classical
boundaries. A significant trend toward integrated, layered
architectures that combine FL, blockchain, and quantum
technologies is also identified. This review concludes that the
evolution of FL in banking is inextricably linked to quantum
advances, demanding a two-pronged proactive strategy: mitigating
immediate risks with PQC while strategically investing in R&D for
QML.