This research aims to reveal the risks resulting from the intensive and unguided use of artificial intelligence applications, including the Chatgpt application, by secondary school students on their cognitive achievement and educational gains. On the one hand, this research aimed to explore the challenges facing the use of artificial intelligence applications in the educational process in Tunisia. On the other hand, the research employed a descriptive-analytical approach based on a theoretical review of the literature related to the research problem and its hypotheses. We also adopted the field method to carry out this research by preparing two research forms that included a set of questions that focused on the reality of employing artificial intelligence applications in the educational and learning process in Tunisia. The first questionnaire was directed at secondary school teachers, while the second was directed at secondary school students. Given the nature of the questions in both questionnaires, we used content analysis to analyze the data, specifically focusing on open-ended questions. This research derives its importance from the value of the results it reached, which confirmed that the excessive and undirected use of artificial intelligence applications by learners was an important factor in the decline of their educational level and the increase in failure rates. The field study of this research also proved that Tunisian educational institutions suffer from the absence of an infrastructure that enables them to effectively employ artificial intelligence applications in order to develop the educational process.