Privacy-Preserving Cybercrime Investigation: AI and Zero-Knowledge Proofs for Secure Network Forensics

Abstract

As cyber-crimes have become more complex network forensics has become an essential element of cybersecurity investigations. However, conventional forensic techniques are confronted with challenges such as data privacy, integrity, and secure authentication of evidence. This paper suggests a privacy-preserving AI-augmented forensic framework that uses Zero-Knowledge Proofs (ZKP) for authenticating forensics securely and blockchain for tamper-evident forensic storage. The intended framework employs AI and ML strategies for real-time intrusion detection real-time intrusion detection, anomaly recognition, and cyber-attack attribution, radically enhancing forensic efficacy and investigative productivity. Experimental evidence obtained with the UNSW-NB15 dataset provides evidence that the AI model offers a detection rate of 97.5% accompanied by precision as high as 96.8% and a recall of as much as 98.2% to ensure good cyber threat classification. Moreover, the verification process of ZKP takes only 1.2 milliseconds, allowing for fast forensic validation with data confidentiality being maintained. The blockchain-based logging system for forensics has an overhead of merely 0.35 MB per transaction, allowing tamper-proof and scalable storage of forensic data. The findings confirm that integrating AI, ZKP, and blockchain improves forensic trustworthiness at the cost of reduced data exposure. This work adds to developing privacy-protecting forensic techniques and offers a secure, scalable solution for contemporary cybercrime investigations.