Abstract
- The integration of forensic audits into artificial intelligence (AI) represents a significant advancement in the field of forensic science. This chapter explores existing methodologies for auditing artificial intelligence, including statistical methods, explainable AI and digital forensic readiness (DFR) procedures, in order to address the deficiency in transparency, especially in continual learning applications. Future perspectives on forensic AI audits, including their legal applicability, the establishment of standardized auditing frameworks, and the necessity for interdisciplinary collaboration to address ethical and legal concerns, are discussed. The findings emphasize the trade-off between the theoretical requirement for DFR in AI and the practical applicability of auditing in real-world scenarios, particularly in safe-critical and time-critical scenarios.