This chapter explores challenges linked to AI integration in professional training, emphasising the need to question the learning analytics generated during activities involving humans or living entities. The intersection of initial and ongoing training poses a design challenge: making AI tools acceptable for both training and the workplace. This chapter critically examines the nature and quality of the initial data and learning analytics using educational data mining through three case studies of adaptive learning environments. The first case study addresses the challenges of modelling Comté cheese manufacturing. The second case study describes Silva Numerica, a digital forest simulator, exploring how AI, as a learning tool, can contribute to realistic modelling while addressing didactic obstacles. The third case delves into AI's role in automotive mechanics training, emphasising the need for visibility in cognitive inference processes. The chapter concludes by addressing data reliability concerns in AI systems and proposing education and training strategies to overcome such challenges.