AI in Physical Security: Opportunities, Risks, and Governance

These sources provide a comprehensive overview of the increasing integration of Artificial Intelligence (AI) and Machine Learning (ML) into the physical and cyber security sectors, examining both the significant opportunities and the associated risks. Several documents explore the conceptual underpinnings of AI, including the different paradigms (Symbolic, Statistical, Sub-Symbolic) and intelligence levels (Narrow, Broad, General), and explain how current security technologies predominantly use Narrow AI techniques for functions like detection and control. A major focus across the texts is the critical need for robust AI risk management and governance, frequently referencing the NIST AI Risk Management Framework (RMF) and regulatory compliance, particularly concerning biometric data privacy laws like GDPR. The texts highlight the benefits of AI in security, such as enhanced efficiency, improved threat detection, and the potential for systems like adaptive access control and thermal AI in perimeter security, while also acknowledging challenges like data quality requirements, the "black box" issue of transparency, and new security vulnerabilities related to AI agents and orchestration.