What Predicts AI Usage? Investigating the Main Drivers of AI Use Intention over Different Contexts

Abstract

Artificial Intelligence (AI) based applications are an ever-expanding field, with an increasing number of sectors deploying this technology. While previous research has focused on trust in AI applications or familiarity as predictors for AI usage, we aim to expand current research by investigating the influence of knowledge as well as AI risk and opportunity perception as possible predictors for AI usage. To this end, we conducted a study (N= 450, representative for age and gender) covering a broad number of domains (health, eldercare, driving, data processing, and art), assessing well-established variables in AI research (trust, familiarity) as well as knowledge about AI and risk and opportunity assessment. We further investigated the influence of AI use related ratings on AI usage. Results show that the newly investigated variables best predict overall intention to use, above and beyond trust and familiarity. Higher AI-related knowledge, more positive use related ratings, and lower risk perception significantly predict general AI use intention, with a similar trend emerging for domain-specific AI use intention. These findings highlight the relevance of knowledge, risk and opportunity assessment, and use related ratings, in understanding laypeople’s intention to use AI-based applications and open a new roster of research questions in understanding people’s AI use behavior intentions and their perception of AI.

Artificial Intelligence in Medicine: The Influence of Medical Expertise and Perceived Causability on Medical AI Risk and Benefit Perception

“…As has been shown in several recent studies, risk and benefit perception are important predictors for AI usage (Potinteu et al, 2023, Schwesig et al, 2023. In the study of the authors identified risk and benefit perceptions as the most important predictor for use intention (with trust in AI only being second place).…”

“…While related research shows the benefit of causability concerning, for example, trust in AI (Druce et al, 2021;Glass et al, 2008;Larasati et al, 2021;Shin, 2021), we are, to our knowledge, the first to relate this concept to risk and benefit perception. Given the importance of risk and benefit perception regarding AI usage (Potinteu et al, 2023, Schwesig et al, 2023, increasing the perceived causability seems to be a promising method to decrease the risk and increase the benefit perception and, thus, increase the probability of AI usage in the medical sector. However, given that being in the high vs. low causability group did not result in any significant differences regarding risk and benefit perception further research in how to increase perceived causability is needed.…”