This paper introduces an AI-powered Knowledge Graph for large forensic data investigations, combining machine learning and deep learning to create a sophisticated digital investigation tool. Traditional forensic methods often suffer from a lack of synergy among experts, leading to missed insights and delayed judicial processes. Our Knowledge Graph addresses this by autonomously identifying connections between offenders or victims and analyzing crime event patterns using machine learning-based knowledge signatures and spatial cascadability metrics. The paper details the creation of a Knowledge Graph from diverse forensic data, highlighting the challenges of data handling and standardization. It showcases the application of this approach in four real-world datasets, demonstrating its effectiveness in forensic reasoning. The results indicate that AI-enabled knowledge graphs can significantly enhance digital investigations. Additionally, the use of spectral analysis for examining realworld interconnections highlights the system’s autonomous capabilities. This AI-driven approach promises more efficient digital investigations and could play a crucial role in reducing security breaches in global businesses.