The National Artificial Intelligence Advisory Committee (NAIAC)- Recommendation: Field Testing of LawEnforcement AI Tools

Abstract

The responsible use of AI in law enforcement requires AI developers to train, test, and audit their AI tools to ensure that the results of a predictive tool are sufficiently accurate, non-discriminatory, rights-respecting, and cost-effective. But the true value and risks of an AI tool will depend on how it operates in the real world. The White House now requires all federal agencies to test an AI tool for performance in real-world settings (OMB Memo M-24-10 §5 (c)(iv)(B)).) Very few resources are available to help guide the AI industry, law enforcement departments, and independent researchers through the process of testing AI tools when they are provisionally used in the field. This report and set of recommendations provide the infrastructure for AI field testing in the context of policing.

Keywords: Artificial Intelligence, Technology Law, Law Enforcement, Criminal Justice, Privacy, AI Bias