Purpose: Physical education (PE) in colleges and universities in the new era focuses on mastering smart PE technology, deepening the connotation of big data education, and cultivating and achieving disciplinary core literacy. The paper designs a monitoring plan based on three key aspects of core literacy, namely, athletic capabilities, health-related actions, and PE morality. It aims at strengthening the value of AI-powered methods in core literacy cultivation, providing more teaching methods to traditional PE classroom teaching, and promoting students’ core literacy. Methods: The paper selects 105 college students from 5 classes who voluntarily take part in the research. It also carries out a 10-week teaching practice, during which students are assigned with homework supported by AI sports software for 33 times and conducts both single-item and general data analysis over 7 items of physical fitness training, 1 competition, and students’ behavior. All the data is provided by the AI software as an auxiliary part to PE and art courses. Results: In the single-item analysis of the physical fitness data comprised of 7 items and 48 times of completion, the final grade of 42 items and times of completion is higher than those of the first-time grade, accounting for 87.5%. There is a substantial increase of the grade of balancing, displacement, and sensitivity-related items, and the maximal fluctuation goes to the grade of coordination items. Besides, single-foot standing (a kind of balancing item), two-person Tetris (a kind of displacement and sensitivity-related item), and dancing (a kind of coordination) are the most popular. As to the general data, the highest degree of class-level participation reaches 84.2%, the lowest degree reaches 68.1%, and the degree of completion, however, is unsatisfactory. There is a substantial fluctuation in the 33 times of completion, showing both rising trend and downtrend to some degree. Besides, most students complete their homework from 21:00 to 3:00 of the next day, accounting for 32.9%, and the general degree of participation of online competition reaches 39.1%. Conclusion: The AI sports software improves the level of physical fitness, a part of athletic capabilities, but needs improvement in strengthening the awareness, habits, and management of individual health of health-related actions, and the ability of keeping forging ahead and going beyond oneself of PE morality.