The Acceptance and Usage of ChatGPT: An Information Adoption Model Perspective

Mark Anthony Camilleri,Adriana Camilleri,Mark Camilleri
DOI: https://doi.org/10.2139/ssrn.4929626
2024-01-01
SSRN Electronic Journal
Abstract:Currently, there are just a few studies that sought to explore aspects influencing the users' engagement with generative artificial intelligence (GenAI) large language models (LLMs) like OpenAI's Chat Generative Pre-Trained Transformer (ChatGPT). To the best of the authors' knowledge, for the time being, there are no contributions that have validated the information adoption model (IAM) to investigate individual perceptions in terms of their confidence levels in the content generated by GenAI algorithms. This research identifies the factors that may induce online users to rely on the information produced by ChatGPT. The methodology integrates IAM constructs, namely, information relevance, information accuracy, source trustworthiness and information usefulness to better understand the extent to which they have an impact on information adoption. Data were gathered from a sample of three hundred twenty-seven (n=327) higher education students. The results from SmartPLS indicated that there were highly significant effects in the proposed structured model. The strongest link was reported between information usefulness and information adoption. However, another noteworthy finding of this study confirmed that information relevance exhibited a substantial influence on information usefulness. In sum, this contribution puts forward a robust conceptual model that is empirically grounded. It sheds light on the users' perceptions about the factors that are influencing their adoption of ChatGPT.
What problem does this paper attempt to address?