Application or functionality? The type of information matters: The impact of information about artificial intelligence on AI usage intention and AI aversion

As AI applications become increasingly ubiquitous in everyday life, there is the question of how much the general public knows about AI technology and whether this knowledge influences behavior. Explaining how an AI application reached a given decision is one approach to enhancing technology acceptance. Usually, the goal of such methods is to deepen users' understanding of specific AI applications and investigate the direct impact on human-AI interaction. Our research takes a broader perspective by asking what kind of information affects people's overall willingness to use AI. Our study investigates whether application-focused or technical information influences AI usage intentions across a wide range of topics. Beyond usage intention, we aim to extend current research by investigating whether such information can reduce AI aversion – a distinct concept complementing other concepts like AI anxiety or technophobia. To this end, we collected N = 513 participants from Germany. Results showed that participants who received information about AI application areas reported a higher AI usage intention and lower AI aversion than the control group. When controlling for age, gender, education, and AI knowledge, both types of information significantly reduced AI aversion. Finally, when running a mediation analysis, results revealed that AI aversion fully mediated the direct relationship between type of information and usage intention. Taken together, our results suggest that communicating real-world use cases may be more effective in enhancing AI usage intention than technical explanations alone. Additionally, our results point towards AI aversion as a potential barrier to technology adoption. Lastly, our results also hint at how easily a limited amount of information can influence usage intention. Thus, future studies should investigate the impact of a more balanced depiction, also accounting for potential pitfalls, on technology adoption.