<p>Artificial Intelligence (AI) tools, especially Large Language Models (LLMs), are rapidly becoming a vital part of the academic lives of students in higher education, changing how they study, complete assignments, and gain support. Despite their importance and prevalence, however, little is known about how the frequency with which students utilize the tools impacts their views on the reliability, academic integrity, and the quality of AI-generated outputs. To address this gap, this study examined the adoption patterns, attitudes, and ethical concerns of 237 university students in the United States. A cross-sectional survey was conducted, and the data was analyzed using descriptive statistics and non-parametric tests: Kruskal–Wallis H, Mann–Whitney U, and Spearman’s rho. The findings showed that while AI tools are widely embedded in students’ academic routines, they are not being used to do the students’ work for them but rather to provide support when they need help. Conversational platforms such as ChatGPT were found to be used by 84.8% of the students surveyed, but significant differences in the frequency of AI use were evident across fields of study, levels of self-rated tech-savviness, exposure to prior AI-related training, and gender. More frequent AI use was reported by students in business, economics, and engineering, those with higher levels of confidence in their technical skills, those who had received AI training; and male students in general. It is also noteworthy that frequent AI use was found to be associated with stronger confidence in the accuracy and quality of AI outputs; fear that it may undermine academic honesty was of less concern. These findings highlight the need for clearer institutional guidance, stronger AI literacy, and more responsible, human-centered integration of AI in higher education.</p>