An increasing number of Artificial Intelligence applications are developed and used in today’s society. Despite this, the public has relatively low knowledge on how AI functions or the extent to which it is deployed. To better understand the relevance of framing AI in terms of how it works vs. where it is deployed, a study (N = 513) was conducted. We compared the influence of brief AI descriptions (functionality framing vs. applications framing vs. control group) on a number of variables: trust in AI, attitudes, risk and opportunity perception, aversion, and AI usage intention. Demographics, knowledge about AI, and confidence in one’s knowledge were assessed. Results showed a significant impact of framing on trust and aversion. Participants who read a description regarding the use of AI (applications) reported higher trust and lower aversion. Results further reveal an interaction effect of age and gender, with younger men placing significantly more trust in AI. Elevated trust in AI was further explained by people’s high confidence but not their knowledge. Lower AI aversion was related to both higher levels of knowledge in AI and higher confidence. Risk and opportunity perception together with attitudes remained similar across the three groups. Additionally, both trust in AI and aversion significantly predict intention to use. Our findings thus highlight the impact AI framing can have on trust and aversion, which in turn relate to people’s intention to use AI applications. These results further bring into question the stability of trust and aversion in relation to AI.