Opportunities and Challenges of Artificial Intelligence in Digital Forensics

Abstract

Digital forensics research remains constrained, while the rapidly evolving digital landscape renders traditional forensic methodologies increasingly inadequate for modern investigative challenges. This work conducts a systematic literature review and bibliometric analysis of computer forensics, specifically targeting digital forensics applications. The study employed a systematic literature evaluation of the Scopus database using "Computer Forensic" as the search term within article titles, abstracts, and keywords. The initial search retrieved 3,222 publications, subsequently refined to 120 academic articles through PRISMA methodology with inclusion criteria encompassing computer science subject areas, final journal articles, English language publications, and open access availability. Three research questions guide this investigation: examining future digital forensic research directions, analyzing current research methodologies, and identifying practical and theoretical implications. Data collection occurred on May 21, 2025, with analysis performed using VOS Viewer bibliometric software. Results reveal that digital forensics research predominantly originates from industrialized nations, particularly the United States and Europe, accounting for 49 of 120 examined articles (40.83%), while Asian and African contributions remain substantially underrepresented. The analysis identified a four-stage digital forensics implementation framework: identification, collection, analysis, and preservation. Furthermore, the investigation examined artificial intelligence applications in digital forensics, particularly NLP-based approaches and machine learning algorithms including CNN models for forensic processes. While AI has revolutionized digital forensics by enhancing accuracy, efficiency, and investigative effectiveness, the analysis reveals persistent challenges including algorithmic bias, data privacy concerns, and decision-making transparency issues. Future research should incorporate additional databases such as Web of Science to enhance data quality and scope. The integration of AI and machine learning models across digital forensics stages promises to deliver more precise and thorough investigative outcomes.