Artificial Intelligence (AI) applications for education are being developed at an increasing pace. It seems reasonable to assume that these applications would enhance student experiences and course satisfaction, and that therefore educational institutions should invest in these technologies to enhance their student offer. However, this should be tested empirically. In the current study a gender-balanced sample of 302 UK students rated course satisfaction, completed the General Attitudes towards AI Scale (GAAIS), comfortableness with AI educational applications, and course satisfaction if AI educational applications were adopted. Although students were, on average, moderately comfortable with AI educational applications, course satisfaction dropped in response to their hypothetical adoption. AI applications that assigned summative grades or that offered wellbeing support gave rise to the highest levels of discomfort. Students were more comfortable with career support, formative course support, and administrative support. Positive and Negative AI attitudes predicted the satisfaction difference, with mediation via comfortableness with applications. We recommend that Higher Education Institutions exercise caution before making major investments in AI educational applications.