Inferring trust in recommendation systems from brain, behavioural, and physiological data

TL;DR

This study investigates how neural, behavioral, and physiological data can objectively measure and understand trust in AI recommendation systems, using music curation as a model, and highlights the potential for multimodal trust calibration.

Contribution

It introduces a multimodal approach combining EEG, pupilometry, and reinforcement learning to objectively assess trust in AI, moving beyond subjective self-reports.

Findings

System accuracy influences user trust and music preferences.

Neural oscillations and pupil responses correlate with reward prediction errors.

Trust calibration can be neurally modeled and potentially improved.

Abstract

As people nowadays increasingly rely on artificial intelligence (AI) to curate information and make decisions, assigning the appropriate amount of trust in automated intelligent systems has become ever more important. However, current measurements of trust in automation still largely rely on self-reports that are subjective and disruptive to the user. Here, we take music recommendation as a model to investigate the neural and cognitive processes underlying trust in automation. We observed that system accuracy was directly related to users' trust and modulated the influence of recommendation cues on music preference. Modelling users' reward encoding process with a reinforcement learning model further revealed that system accuracy, expected reward, and prediction error were related to oscillatory neural activity recorded via EEG and changes in pupil diameter. Our results provide a neurally grounded account of calibrating trust in automation and highlight the promises of a multimodal approach towards developing trustable AI systems.