Remote AI Weapons Detection – reversing stop and search to search and stop?

This paper critically examines the legal and normative implications of remote artificial intelligence (AI) weapons detection systems currently under trial in the United Kingdom. Positioned as a technological solution to rising knife crime and counter-terrorism challenges, these systems claim to reduce reliance on traditional stop and search practices by identifying concealed weapons without physical contact. However, we argue that such technologies invert the established logic of stop and search – transforming it into a form of “search and stop” – with significant consequences for individual rights (under the ECHR) and police accountability. Through a doctrinal analysis of the Police and Criminal Evidence Act 1984 (PACE), relevant case law, and the surveillance regime under the Regulation of Investigatory Powers Act 2000, the paper explores whether remote scanning constitutes a search in law, whether its use triggers procedural safeguards under section 2 of PACE, and whether AI-generated outputs can satisfy the requirement for “reasonable grounds”. The paper further interrogates the use of suspicion-less search powers under section 60 of the Criminal Justice and Public Order Act 1994 (as amended) and evaluates the compatibility of remote detection with the right to privacy, equality law, and the Peelian principle of policing by consent. Drawing on recent controversies surrounding similar systems deployed in countries such as the United States, and informed by developments in human rights and public law, we conclude that remote weapons detection cannot be justified under existing legal frameworks without reform. In light of this, the paper recommends urgent regulatory action, including legal clarification, independent oversight, and equality impact assessments, to ensure that innovation in policing does not come at the expense of legality, transparency, and public trust.