JMIR Publications

ABSTRACT

Background:

ChatGPT, a language model developed by OpenAI, has gained immense popularity among computer users owing to its remarkable ability to generate responses that closely resemble those of human language. It is essential to note that overreliance or blind trust in ChatGPT, especially in high-stakes decision-making contexts, can have severe consequences. Similarly, lacking trust in the technology can lead to underutilization, resulting in missed opportunities.

Objective:

This study aims to capture users' trust in ChatGPT and explore its impact on users’ intent and actual use of the technology. The following hypotheses were tested: H1: User intent to use ChatGPT increases with their trust in the technology. H2: The actual use of ChatGPT increases with users’ intent to use the technology. H3: The actual use of ChatGPT increases with users’ trust in the technology. H4: Users’ intent to use ChatGPT can partially mediate the effect of trust in the technology on its actual use.

Methods:

This study obtained ethical approval and distributed an online survey to adults in the US who actively use ChatGPT at least once a month to examine the relationship between trust, intent to use, and actual use of ChatGPT for healthcare queries. The survey was soft launched with 40 responses and then distributed to a larger audience, collecting data from February 2023 through March 2023. Participants responded to survey questions on a Likert scale, and two latent constructs were developed: Trust and Intent to Use, with Actual Use being the outcome variable. Descriptive statistics of study variables were calculated. The study used SEMinR package and PLS-SEM method for multivariate analysis to evaluate and validate the latent constructs' convergent and discriminant validity and test the structural model and hypotheses. PLS-SEM was deemed suitable for exploratory research as it offers the flexibility needed for the interplay between theory and data.

Results:

In the study, 607 respondents completed the survey. Among them, 182 (30%) used ChatGPT at least once a month, with 158 (26%) using it once per week, 149 (25%) using it more than once per week, and 118 (19%) using it almost every day. A majority of respondents had a minimum high school diploma (n=204, 34%) or a bachelor's degree (n=262, 43%). The primary uses of ChatGPT were for information gathering (n=219, 36%), entertainment (n=203, 33%), and problem-solving (n=135, 22%), with a smaller number using it for health-related queries (n=44, 7%) and other activities (n=6, 1%). The model explained 50.5% and 9.8% of the variance in "Intent to Use" and "Actual Use," respectively, with path coefficients of 0.711 and 0.221 for "Trust" on "Intent to Use" and "Actual Use," respectively. The bootstrapped results failed to reject all four hypotheses, with trust having a significant direct effect on both intentions to use (ß = 0.711, 95% CI [0.656, 0.764]) and actual use (ß = 0.302, 95% CI [0.229, 0.374]). The indirect effect of trust on actual use, partially mediated by intention to use, was also significant (B = 0.113, 95% CI [-0.003, 0.227]).

Conclusions:

The study provides novel insights into the factors driving the adoption of chatbot technologies such as ChatGPT. Our results suggest that trust is critical to users' adoption of ChatGPT. Companies and policymakers should prioritize building trust and transparency in developing and deploying chatbots. AI has undergone a transformative evolution since its inception in the 1950s. The perception of AI has evolved over time, from a magical panacea to all problems to a technology that can augment human capabilities and solve complex problems. With the rapid advancements in AI research, it is evident that AI will continue to revolutionize various fields and shape the future of technology.