AI-Based Biometric Surveillance in Public Spaces: Balancing Privacy Rights and National Security

Abstract

Biometric surveillance is used in public spaces for identification purposes, as well as for public and state security, but raises concerns about legality, privacy, mass surveillance, and errors. This study constructs a Legal–Technical Proportionality Matrix to evaluate whether the use of artificial intelligence for biometric surveillance is legally-technically proportionate. A doctrinal-comparative analysis was conducted to examine the relevant privacy, data-protection, and AI governance frameworks. These were tested against technical performance assessments of facial-recognition technology that demonstrated varying degrees of accuracy and applicability. Four scenarios were evaluated using the matrix, resulting in compatibility scores of 90% for a time-bound survey of a missing person, 85% for a focused screening of a watch-list at an airport, 30% for a city-wide constant surveillance and 25% for identifying peaceful protestors. It was concluded that technical accuracy alone cannot overcome legal limitations, and that biometric surveillance should be implemented as a targeted, independently authorized, technically reliable, transparent, time-boxed, and remediable measure.