Artificial intelligence (AI) has rapidly expanded across medicine, demonstrating value in image analysis, risk prediction, and data interpretation. In forensic science, AI has been explored or implemented in areas such as fingerprint examination, facial recognition, and ballistics; applications specific to medicolegal death investigation and forensic pathology (FP) remain largely experimental. This review summarizes current and emerging research on AI in FP, including studies focused on gunshot wound classification, postmortem interval estimation, toxicological analysis, cause and manner of death determination, and postmortem imaging. Several proof-of-concept investigations demonstrate promising performance, though most rely on limited data sets and lack external validation. Despite its potential to enhance efficiency, consistency, and reproducibility, widespread adoption of AI in FP is constrained by significant challenges. These include limited data access, potential model bias, limited generalizability, difficulties in validation, ethical concerns surrounding decedent data, and medicolegal concerns related to transparency and admissibility in court. AI-related governance frameworks offer guidance for mitigating these risks, but will need to be adapted to the FP context. Overall, AI should be viewed as a complementary tool rather than a replacement for forensic pathologists. Thoughtful development, rigorous validation, transparent implementation, and robust oversight will be essential for integrating AI into FP in a scientifically sound and legally defensible manner.
Atherton et al. (Wed,) studied this question.