Trust in Facial Recognition Systems: A Perspective from the Users.

High-risk artificial intelligence (AI) are systems that can endanger the fundamental rights of individuals. Due to their complex characteristics, users often wrongly perceive their risks, trusting too little or too much. To further understand trust from the users’ perspective, we investigate what factors affect their propensity to trust Facial Recognition Systems (FRS), a high-risk AI, in Mozambique. The study uses mixed methods, with a survey (N = 120) and semi-structured interviews (N = 13). The results indicate that users’ perceptions of the FRS’ robustness and principles of use affect their propensity to trust it. This relationship is moderated by external issues and how the system attributes are communicated. The findings from this study shed light on aspects that should be addressed when developing AI systems to ensure adequate levels of trust.