Article

Machine Learning-Based Classification of Cyber Attacks and Its Legal Implications for Cybercrime Enforcement 

---

Abstract

The increasing sophistication and frequency of cyber-attacks present significant challenges for both technical detection and legal enforcement. This study aims to classify various types of cyber attacks using a machine learning approach and to examine the legal implications of such classification for cybercrime enforcement. A dataset consisting of 40,000 network traffic records with 25 technical and contextual features was analyzed using the Random Forest algorithm. The model achieved an overall accuracy of 33.8%, with balanced precision and recall across three major categories: Distributed Denial of Service (DDoS), Malware, and Intrusion. The feature importance analysis revealed that Anomaly Scores and Packet Length were the most influential predictors in detecting malicious activity, suggesting that behavioral and quantitative network indicators are more effective for identifying threats than static categorical variables. These technical findings hold important legal significance, as quantifiable digital metrics can strengthen the reliability, transparency, and admissibility of forensic evidence in court proceedings. Furthermore, the balanced distribution of attack types highlights the need for comprehensive and adaptive cybercrime legislation that accommodates multiple threat categories and keeps pace with rapid technological developments. Aligning national laws with international frameworks such as the Budapest Convention on Cybercrime is essential for ensuring technological neutrality and effective prosecution of digital offenses. Overall, the study bridges the gap between machine learning and legal analysis by demonstrating that artificial intelligence can serve not only as a technical tool for cyber threat detection but also as a foundational element for evidence-based and accountable cyber law enforcement.

Keywords: Cybersecurity, Machine Learning, Digital Forensics, Cybercrime, Cyber Law

How to Cite:

Natania, D., (2026) “Machine Learning-Based Classification of Cyber Attacks and Its Legal Implications for Cybercrime Enforcement ”, Journal of Cyber Law 2(1). doi: https://doi.org//JCL.150