The influence of real-time trust assessment and adaptive feedback on users' trust in Artificial intelligence: Autonomous vehicle case study

Research into human-system symbiosis has increasingly focused on real-time trust assessment using psychophysiological signals.While previous studies have developed and evaluated machine learning-based cognitive models of trust, none have implemented these models in real time with integrated audio-visual feedback as a trust-reinforcing mechanism.Moreover, no studies have explored whether such real-time assessments, combined with adaptive feedback, can enhance user experience in terms of trust and reliance.This study addresses that gap by evaluating how an adaptive feedback system for measuring users' trust in real-time improves users' experience (trust and overall reliance).The results of a within-subject experiment involving four risk-level conditions in an autonomous vehicle driving simulation show that real-time trust assessment with adaptive feedback significantly enhances user trust and reliance on AI-driven systems.