Artificial intelligence (AI) technology has taken a firm place in engineering education, creating illusions of improving the quality of mastering engineering knowledge at universities. Students and teachers are enthusiastically coming up with ways to improve the efficiency of their work. The use of artificial intelligence causes heated debates in the scientific literature, noting conflicting opinions on opportunities and risks for improving the quality of education. This contribution presents a study that, along with the results of generalizing the scientific and analytical literature, using user experience, investigates the point of view of ChatGPT on the problems and consequences of the spread of AI in engineering education. The results show that the growing trust in AI among students and teachers is accompanied by illusions of obtaining deep knowledge and that their work is independent. There are illusions about the individualization of the learning trajectory, objectivity and completeness of information, the sufficiency of «virtual practice». AI does not yet have the ability to objectively assess the degree of independence in performing work. The conclusion of the study emphasizes the significance of timely intervention in the use of AI at universities in the training of engineering personnel. The obtained results can serve as a guide for developing guidelines for the application of AI at universities and for future research aimed at understanding the significance of AI in education.