• Vinutha GR
• Dr.Nirmala Hiremani

Abstract—Cyber threats targeting healthcare and financial institutions have grown sharply with the digitization of critical services, yet effective detection increasingly depends on intelligence shared across organizations—intelligence that privacy regulations such as HIPAA, GDPR, and PCI-DSS forbid these organizations from pooling. Federated Learning (FL) resolves this tension by enabling collaborative model training in which only model updates, never raw data, are exchanged. This paper presents a comprehensive survey of FL-based threat detection and proposes the Privacy-Preserving Federated Threat Intelligence Framework (PPFTIF) for healthcare and financial ecosystems. PPFTIF integrates federated learning, differential privacy, secure aggregation, Byzantine-resilient aggregation, and Explainable AI (XAI) to achieve collaborative threat detection while preserving data confidentiality. We formalize a sector-weighted federated objective, specify the end-to-end training algorithm with its privacy and robustness steps, and analyze the framework against membership-inference, gradient-leakage, and model-poisoning attacks. A worked cross-sector example shows how a hospital and a bank jointly improve detection of a shared attack campaign without exposing any record, and a layered evaluation methodology with explicit datasets, metrics, and a privacy-budget protocol is provided. The contribution is an integrated, explainable, and regulation-aware blueprint rather than empirically validated outcomes; the design directly addresses cross-sector collaboration, poisoning resilience, explainability, and simultaneous regulatory compliance.

"Federated Learning for Privacy-Preserving Threat Detection: Collaborative Attack Detection Without Sharing Data in Healthcare and Finance", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 10, page no.c866-c871, October-2026, Available :http://www.jetir.org/papers/JETIR2610287.pdf

"Federated Learning for Privacy-Preserving Threat Detection: Collaborative Attack Detection Without Sharing Data in Healthcare and Finance", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 10, page no. ppc866-c871, October-2026, Available at : http://www.jetir.org/papers/JETIR2610287.pdf