Epistemological Beliefs as Predictors of Generative AI Familiarity, Perceived Issues Likelihood, and Usage

With the rising popularity and ethical concerns about generative artificial intelligence (GAI), there are strong interests in understanding the factors behind its perception and use. Epistemological beliefs, found pertinent to information behavior, are rarely studied in GAI usage research. This study conducted path analysis on survey responses from 322 U.S. adults to explore how epistemological beliefs (from Schommer's Epistemological Questionnaire) relate to GAI perception (familiarity and perceived likelihood of GAI issues) and use (frequency and seriousness). The study found direct positive paths from Avoid Ambiguity beliefs to perceived GAI issues likelihood and usage frequency, and indirect positive paths from Depend on Authority to frequency and seriousness of use via familiarity with GAI. Theoretical implications, and practical implications for information literacy, are discussed.