MLPhishChain: a machine learning-based blockchain framework for reducing phishing threats

Introduction Phishing attacks pose a significant threat to online security by deceiving users into divulging sensitive information through fraudulent websites. Traditional anti-phishing approaches are centralized and reactive, exhibiting critical limitations such as delayed detection, poor adaptability to evolving threats, susceptibility to data tampering, and lack of transparency. Methods This paper presents MLPhishChain, a decentralized application (DApp) that integrates blockchain technology with machine learning (ML) to provide a proactive and transparent solution for URL verification. Users can submit URLs for automated phishing analysis via an ML model, with each URL’s status securely recorded on an immutable blockchain ledger. To address the dynamic nature of phishing threats, MLPhishChain features a re-evaluation mechanism, enabling users to request updated assessments as URLs and website content evolve. Additionally, the system incorporates data from external security services (e.g., VirusTotal) to offer a multi-source validation of phishing risk, enhancing user confidence and decision-making. Results The system was built using Ganache and Truffle, and performance metrics were computed to evaluate its efficacy in terms of latency, scalability, and resource consumption. Results indicate that the proposed system achieves rapid URL verification with low latency, scales effectively to handle increasing user submissions, and optimizes resource usage. Discussion By leveraging the strengths of decentralized blockchain technology and intelligent ML algorithms, MLPhishChain addresses the shortcomings of traditional anti-phishing methods. It delivers a reliable and adaptable solution capable of addressing the evolving nature of phishing threats. This approach establishes a new standard in phishing detection, characterized by enhanced transparency, resilience, and adaptability.