AI-enhanced security framework for blockchain

Blockchain is a decentralized, distributed ledger technology that securely records transactions across multiple nodes, ensuring transparency and immutability. It has the potential to significantly reduce costs and time while minimizing risks across various applications. This study presents an AI-enhanced approach to strengthen blockchain security, with a focus on transaction monitoring and anomaly detection. Challenges in transactions include fraudulent activities, double-spending, and unauthorized access, while anomalies often signify errors, potential hacking attempts, or irregular patterns that may undermine the integrity and trust of blockchain systems. By implementing a self-triggering mechanism that utilizes AI-computed features, the system monitors transactional behaviour in real-time to detect and mitigate potential cyber-attacks. Using Isolation Forest algorithm for anomaly detection, the system effectively tracked and identified unusual patterns within a five-node star topology blockchain network. Upon detecting anomalies, the system autonomously removes compromised nodes to prevent further damage and maintain operational integrity. Simulated results show that the system effectively identified compromised nodes without disrupting business transactions, maintaining an average transaction amount of 27.50 units across six transactions. Using AI with blockchain offers a scalable solution that enhances both the security and efficiency of blockchain networks compared to traditional, rule-based or manual approaches.