Navigating the Evolving Landscape of Human Reliance on Generative AI

Abstract

The rapid diffusion of generative artificial intelligence into workplace and personal contexts has created substantial uncertainty regarding how individuals rely on these systems over time. This article examines longitudinal patterns in AI delegation and disclosure behaviors, drawing on evidence from a six-wave study spanning ten months with over 1,000 U.S. participants. Contrary to expectations of increasing familiarity breeding greater reliance, findings reveal declining willingness to delegate tasks and disclose information to AI systems, particularly for personally meaningful activities. Professional writing contexts maintained relatively stable delegation patterns, while personal contexts showed pronounced declines. Trust, anthropomorphic perceptions, and positive attitudes toward AI emerged as consistent predictors of both delegation and disclosure behaviors. These patterns suggest that human-AI interaction reflects a calibration process rather than simple habituation, with important implications for organizational AI adoption strategies, training programs, and technology design. Organizations must recognize that sustainable AI integration depends less on repeated exposure and more on cultivating well-calibrated trust, contextually appropriate applications, and meaningful transparency mechanisms.