TRUST CALIBRATION IN HUMAN AI SURVEILLANCE TEAMS SUPPORTED BY A PSYCHOMETRIC FRAMEWORK FOR HR DECISIONS

Abstract

trust calibration in human-AI surveillance teams, but it is very dependent on calibrated trust between the human agent and the AI system. Mis-calibrated trust, such as hyper-reliance or hyper-scepticism, can degrade the quality of decisions, response times, and overall effectiveness of those teams in a mission. This paper develops a psychometric framework to support trust calibration in human-AI surveillance teams based on psychological theory and organizational decision-making models. Based on the TPMAP (Trust, Personality, Mission Alignment, and Performance) framework, we utilize cognitive and affective psychometric indicators (e.g., risk perceptions, personality traits, mood) to illuminate and propose ways to predict each individual's trust threshold. Our research will be applied to assist HR decision-making (e.g., selecting the right cadre of employees, establishing adaptive supervision and in-depth assessments of high consequence/dynamic risk interactions between surveillance personnel) which is why we include psychometric assessments for trust during training selection, supervisory oversight, and risk detection and evaluation of high consequence surveillance conditions in situations where bad actors threaten national interest, public safety, or developing regional catastrophe [5]. By matching teams and tasks to the psychometrically matched trust profile of the team, this process will lead to adaptive calibration mechanisms from contexts suited to match human and AI to manage functioning and trust under varied operations. This model will optimally improve the cohesiveness of the unit and resilience of the surveillers, coupled with transparency, accountability, and ethical compliance in an AI-enhanced or co-support surveillance system. This research links psychological profiling with human-AI system engineering while providing strategic avenues for HR departments and the workplace more broadly to increase performance, manage operational risk, and foster a culture of psychologically informed AI integration.