Investigation of large language models, GenAI, and proprietary AI systems: Digital forensic evidence, readiness and regulation

Contribution Summary

The paper "Investigation of large language models, GenAI, and proprietary AI systems: Digital forensic evidence, readiness and regulation" explores the challenges of investigating AI systems in digital forensics. As AI models become increasingly complex and proprietary, it is crucial to develop methods for preserving and examining digital evidence within these systems. The author highlights the need for AI forensic readiness, including the ability to preserve and disclose relevant data, such as model parameters, system instructions, and user interactions. The paper also discusses the regulatory context, including the EU AI Act and NIS2 Directive, which impose obligations on AI providers to maintain technical documentation and logs. The author argues that digital forensics must adapt to the unique challenges of AI investigations, including the need for scoped, privacy-aware, and legally controlled preservation and disclosure of evidence. With the DOI 10.1016/j.fsidi.2026.302135, this research contributes to the development of AI forensics as a distinct domain of inquiry.

Keywords: AI forensics; digital forensics; large language models; GenAI; proprietary AI systems; regulatory compliance; cybersecurity; investigative techniques