Abstract
Purpose: This paper examines how social presence generated by Generative AI chatbots in organizational settings fosters interpersonal trust and overtrust, and explores implications for employees’ cognitive agency. Methodology: Drawing on Media Equation Theory and the Computers/Chatbots as Social Actors perspective, the paper develops a conceptual framework through a critical synthesis of literature on social presence, trust calibration, and cognitive offloading. Findings: Miscalibrated trust – overtrust - may lead employees to delegate decision-making, reasoning, and evaluation to GenAI chatbots, reducing critical thinking, weakening reflective autonomy, and progressively eroding organizational knowledge and learning capacity over time. Managerial implications: Organizations should promote mindful and reflective GenAI use, positioning chatbots as cognitive collaborators that augment rather than replace human reasoning, and implement governance practices and training initiatives that support appropriate trust calibration. Research limitations: As a conceptual study, the framework requires empirical validation. Future research should use qualitative methods and examine differences across industries, tasks, and cultural contexts. Originality: The paper integrates social presence, trust calibration, and cognitive offloading into a unified framework explaining how GenAI may reshape human cognition, trust dynamics, and knowledge processes in organizational environments.