This article examines the integration of generative artificial intelligence (AI) into the teaching of French for Specific Purposes (FOS), with a particular focus on veterinary medicine. While research on AI in language education is rapidly expanding, studies specifically addressing FOS remain scarce. Our contribution relies on a dual methodology: the design of three pedagogical sequences for a veterinary French textbook, and inter-views with four co-authors involved in this international project supported by the AUF.The analysis highlights several benefits of AI. It facilitates the creation of materials aligned with the CEFR, quickly generates standardized texts (CVs, cover letters, reports), provides varied exercises, and enables the simulation of contextualized professional situ-ations. These tools are considerable time-savers for teachers that foster learner autonomy, particularly through action-oriented and interactive tasks.However, the study also underlines important limitations. Some generated content proves redundant or imprecise, terminological reliability requires careful checking, and the authenticity of AI-generated texts remains a challenge. Human expertise is therefore indispensable, for both adapting materials and ensuring their pedagogical and discipli-nary validity. These findings stress the central role of teacher reflexivity and critical eval-uation when integrating AI into language education.In conclusion, AI should be regarded as a pedagogical assistant rather than a substitute, offering promising perspectives for veterinary FOS. This reflective case study emphasizes the need to develop critical AI literacy among both teachers and students, and calls for further experimentation, particularly through evaluating the textbook with learners.

Rights