As digital investigations become increasingly complex, traditional manual forensic tools struggle to scale effectively. The vast volume of data and the interconnectedness of cases make adding human resources impractical. Recent advancements in Large Language Models (LLMs), such as Microsoft’s Co-Pilot, offer promising solutions by enabling coding and workflow automation, thereby enhancing the efficiency and affordability of digital forensics.

Despite their success in software development, LLMs have not been systematically adapted to the specific challenges of digital forensics, particularly in automating workflows and interpreting results from forensic analysis tools. This gap presents an opportunity to revolutionize digital forensic investigations.

The DIFPILOT project aims to bridge this gap by developing and evaluating a novel approach to assist digital forensic investigators in automating workflows and analyzing data using LLMs. The project will

1. systematically study and assess existing forensic processes in two specific use cases,
2. develop algorithmic methods and models to support the creation and interpretation of forensic workflows, and
3. establish a controlled laboratory environment to evaluate the efficiency and effectiveness of the proposed solutions.

The novelty of DIFPILOT lies in its targeted application of LLMs to the unique challenges of digital forensics, offering both technical innovation and practical relevance. Research outcomes will be disseminated through high-quality scientific publications and prepared for commercial exploitation by the project’s industry partner within 3–5 years of completion.