Despite the increasing use of artificial intelligence as a support tool in education, many college students face challenges when tackling mathematics, even when employing advanced technologies such as ChatGPT to improve their problem-solving skills. The present study aims to analyze college students’ perception of using ChatGPT to solve mathematical exercises and assess how this perception relates to their level of knowledge about artificial intelligence. The methodology, which included comparing scales of perception and knowledge of artificial intelligence between different groups of students, was designed for practicality and efficiency. Surveys were applied in three courses for the first level of university students to collect data. The results showed no significant difference in the perception and knowledge of artificial intelligence between the groups of students, regardless of the semester level, with \(\alpha =0.05\) and \(p-value=0.62\) . In addition, a moderate correlation was found between the knowledge scale of artificial intelligence and the perception of its usefulness in third-semester students. This suggests that, although the level of AI knowledge does not vary significantly between students at different levels, those with more excellent knowledge tend to value the usefulness of tools such as ChatGPT more. The main conclusion of the research is that the use of scales of knowledge, perception, and application of AI allows for a practical analysis of the data collected, fulfilling the study’s objectives. It is suggested that, to improve the perception of the usefulness of AI in mathematics, it is necessary to deepen the training of students in AI technologies, integrating this knowledge from the first educational levels.