Human Trust in Artificial Intelligence: Review of Empirical Research
Ella Glikson, Anita Williams Woolley
Academy of Management Annals · 2020
Abstract
A wide-ranging review that organizes the empirical evidence on when and why people trust AI, distinguishing the cognitive and emotional foundations of trust and how they vary with the form the AI takes.
Methodology
Systematic review of empirical studies on human–AI trust, structured around AI representation (robotic, virtual, embedded) and the tangibility, transparency, reliability, and immediacy of the system's behavior.
Findings
Trust in AI has distinct cognitive (competence, reliability, transparency) and emotional (tangibility, immediacy behaviors, anthropomorphism) drivers; more tangible and transparent systems earn cognitive trust more readily, while emotional trust depends on human-like cues. Trust is dynamic and calibrated differently across representations of AI.
Applied nudge
Build trust on the dimension that matters for the task: emphasize transparency and reliability for high-stakes cognitive trust, and human-like, responsive cues where emotional comfort drives adoption.
Citation
Glikson, E., & Woolley, A. W. (2020). Human trust in artificial intelligence: Review of empirical research. Academy of Management Annals, 14(2), 627-660.
↗ Download PDF