Artificial intelligence has rapidly and surprisingly made its way into the world of education, transforming learning methods. Fundamental questions are being raised about the role of teachers, the evolution of their professions and the way AI can be used, its opportunities and challenges, challenging them to rethink their role. Will the teaching profession undergo a gradual evolution or a radical transformation ? Beyond the media hype surrounding AI, we seek to understand the tangible impact of the use of generative AI in education. What uses are actually being developed in the field? Under what conditions are these new tools transforming learning practices?

The rapid adoption by young people of generative AIs, and ChatGPT in particular, has really worried the educational community. For teaching teams, this massive adoption represents a source of concern and questioning: how can academic integrity be guaranteed ? What role is left to the teacher if the AI can respond faster and sometimes better than we can ? These concerns are legitimate and well-founded, but they should not overshadow an important reality: education has always evolved in response to technological advances.

While many works explore the theoretical potential of this artificial intelligence, our research is distinguished by an empirical approach focused on the real, concrete uses of higher education teachers. Our study is based on thirty interviews conducted with teachers over the course of the academic year. The results show that teachers are unanimously aware of the need to develop their profession. AI is seen as a facilitating tool to help create courses and automate specific tasks (emails, spelling and syntax correction, translation), thus relieving teachers of energy- and time-consuming activities. Thanks to these tools, teachers have increasingly moved away from their primary role of transmitting knowledge towards other missions: that of guiding, accompanying and developing critical thinking. Unbeknownst to him or her, the trainer has gradually broadened his or her spectrum of know-how, drawing up a new professional profile: a designer and architect of educational content, a coach of students' personal development, a developer of cross-disciplinary skills, an ethical guarantor and controller.

Our results also underline the importance of integrating generative AI into learning as a pedagogical tool in its own right, and not as a compensatory tool for students. It's a question of moving from a logic of knowledge reproduction to one of creation, adaptation and intellectual reflexivity while using AI. The educational challenge is shifting from the passive transmission of knowledge to the transmission of meta-cognitive skills: knowing how to navigate, search, evaluate, select the right information and give meaning to knowledge in an overabundant information environment. Teachers must therefore focus on socio-cognitive skills (critical thinking, taking a step back, creativity), for which AI cannot be relevant. This complementarity between human subjectivity and machine objectivity will ensure better quality learning.

The arrival of generative AI in general, and ChatGPT in particular, should not be seen as a challenge to the competence and profession of teachers, but as an invitation to redefine what teachers expect from students, and to reflect on new learning and assessment methods.

Keywords: Artificial Intelligence, Education, Teaching, Transformation, Learning, Pedagogy, Innovation.