Tracking user trust and mental states during cyber-attacks: A survey of existing methods and future research directions on AI-enabled decision-making for the Royal Canadian Navy

G. Robert Arrabito, Ming Hou, Sebastian Fischmeister, Tiago H. Falk, Hannah Willoughby, Madison Cameron, Liam Foley, Sarah Normandin, Simon P. Banbury

Abstract: The increased use of artificial intelligence to detect cyber-attacks against military organizations can put the operator at risk of having miscalibrated trust in the autonomous system. The present paper reviews techniques to capture the human experience of cyber-attacks to support advanced cybersecurity training strategies for the Royal Canadian Navy. Psychophysiological sensors can provide objective measures to help predict the changing mental states of cyber operators (e.g., predict cognitive load, trust, stress) of whether the operator is aware of ongoing human-noticeable aspects of cyber-attacks. Techniques for measuring the experience of cyber-attacks will help advance research on decision-making in automated systems.