Integrating Artificial Intelligence and Blockchain for Secure Digital Evidence Management: Advances in Cyber Forensics and Data Privacy
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
This research offers a quantitative assessment of a combined AI–Blockchain–Quantum forensic system aimed at improving the accuracy, integrity, and confidentiality of managing digital evidence. The research utilized three extensive datasets, each containing 5,000 entries for deepfake detection, blockchain chain-of-custody, and cybercrime profiling, and conducted experimental simulations to evaluate the effectiveness of sophisticated computational models. In the deepfake detection study, Transformer-based models obtained a mean accuracy of 96.8%, precision of 95.4%, recall of 97.1%, and an F1-score of 96.2%, surpassing CNN (91.3%) and GAN (88.6%) models. The Receiver Operating Characteristic (ROC) analysis produced an AUC value of 0.982, confirming excellent classification performance. Forensic tracking using blockchain technology showed a verification success rate of 98.4% and an average transaction delay of 1.62 seconds over 2,000 recorded transactions, emphasizing the system's scalability and resistance to tampering. Simulations of quantum encryption showed a 37% drop in privacy risk index and a 64% reduction in decryption success rates for classical algorithms under quantum attacks. In cybercrime profiling, models driven by AI like Random Forest, SVM, and LSTM attained classification accuracies of 91.7%, 89.4%, and 94.2%, respectively, across three main categories: financial fraud, phishing, and data exfiltration. Statistical assessments employing confusion matrices, correlation analyses, and ROC evaluations verified the dependability of the framework. Together, these findings illustrate that combining AI analytics, blockchain'sunchangeability, and quantum-resistant cryptography can boost forensic precision by more than 25%, enhance chain-of-custody clarity by 30%, and minimize privacy risks by almost 40%, positioning the suggested system as a forward-looking framework for digital forensic inquiries and legal acceptance.