Explainable AI in High-Stakes Forensic Decision-Making
Hermano Jorge Da Silva De Queiroz · Advances in computational intelligence and robotics book series · 2025
In high-stakes forensic contexts—such as criminal investigations, counter-terrorism, and judicial processes—the use of artificial intelligence (AI) has expanded significantly. Yet, the opacity of many AI models, especially deep learning systems, raises serious concerns about trust, accountability, and legal admissibility. Explainable Artificial Intelligence (XAI) has become essential for enhancing transparency and interpretability in forensic environments. This chapter explores the theoretical foundations, leading methods, and real-world applications of XAI in digital forensics. It also addresses the ethical and legal implications of integrating XAI into forensic workflows and evaluates frameworks for balancing performance and interpretability. Case studies from law enforcement and cybersecurity highlight both the practical advantages and limitations of explainable models. Ultimately, the chapter advocates for a principled, context-aware approach to ensure AI-assisted decisions remain both accurate and understandable to human stakeholders.