Arti­fi­cial intel­li­gence (AI) has become an indis­pens­able tool in many fields, includ­ing health­care, finance, trans­port and logis­tics and also in indus­try. It offers inno­v­a­tive solu­tions for improv­ing effi­cien­cy, reduc­ing costs and opti­miz­ing process­es. UTC, for its part, is involved in a num­ber of AI tool devel­op­ment projects, both in col­lab­o­ra­tion with exter­nal part­ners such as the Sor­bonne Uni­ver­si­ty Alliance as part of the PostGenAI@ Paris IA-Clus­ter, and with­in the uni­ver­si­ty’s own lab­o­ra­to­ries. Some, like LMAC (Applied Maths) and the CID (Knowl­edge, Uncer­tain­ty and Data) team at UTC-Heudi­asyc, are work­ing on AI mod­els them­selves, while oth­ers are work­ing on spe­cif­ic appli­ca­tions linked to their research projects. How­ev­er, while the ben­e­fits of AI can be sig­nif­i­cant, the risks of error are unac­cept­ably high in many fields. Hence the adop­tion of the Euro­pean AI Act by the Euro­pean Union (EU).