Artificial intelligence (AI) has rapidly emerged as a transformative force in higher education, prompting a reevaluation of longstanding academic practices. Although AI tools such as ChatGPT and adaptive learning systems are increasingly being integrated into university courses, the true extent of their impact on student learning remains under investigation. This paper presents a case study conducted in the spring semester of 2024, examining the influence of AI usage on students’ academic performance and learning behaviors in an asynchronous course, as well as the effects of other learning behaviors on academic performance. Students were randomly assigned to two groups, those permitted to use AI tools and those restricted (by having signed a legal contract) from doing so, enabling a comparative analysis of their grades, engagement patterns, and reliance on course materials. Contrary to widespread expectations, no statistically significant difference in final grades was observed between the AI and non-AI groups, suggesting that AI tools alone do not inherently enhance academic outcomes. However, consistent across both groups was a strong correlation between active engagement during the course and final academic performance, emphasizing the importance of learning behaviors over tool usage. These findings contribute to the growing discourse on the role of AI in higher education, raising critical questions about when, how, and for whom AI integration in academic settings is most effective.