The episode examines the impact of GPT-4 on university exams, highlighting how artificial intelligence has managed to generate high-quality answers, often surpassing those of students, without being detected by evaluators. This raises significant concerns about academic integrity and underscores the need to reconsider traditional assessment methods. The study proposes rethinking educational approaches by integrating AI as a teaching tool and focusing on skills that AI cannot replicate. Finally, it emphasizes the urgency of developing more resilient evaluation systems to prevent the unauthorized use of artificial intelligence.