Abstract: The integration of blockchain technology with artificial intelligence (AI) presents a transformative approach to enhancing cybersecurity systems. This paper proposes a comprehensive framework combining decentralized AI models with blockchain’s immutable ledger capabilities to create robust security solutions. Our methodology employs federated learning for privacy-preserving threat detection while utilizing smart contracts for automated response mechanisms. Through extensive experiments on a dataset of 150,000 cyber threat samples across 25 attack categories, we demonstrate a 98.7% detection accuracy with 45% reduction in false positives compared to centralized systems. The implemented system shows particular effectiveness against advanced persistent threats (APTs) and zero-day attacks, achieving 97.1% recall for previously unseen threats. We develop a practical deployment architecture suitable for enterprise environments with throughput of 2,500 transactions per second, and conduct real-world validation with five industry partners. This work contributes significant advances to the field of decentralized cybersecurity by providing a scalable, tamper-proof solution that maintains data privacy while improving threat intelligence sharing among organizations, along with detailed performance benchmarks across multiple deployment scenarios.

Index Terms: Blockchain, Artificial Intelligence, Cybersecurity, Federated Learning, Smart Contracts, Threat Detection, Decentralized Systems, Privacy Preservation

| DOI: 10.17148/IARJSET.2025.124100