The adoption of AI in higher education is considered promising for enhancing the inclusivity of learning trajectories for neuro-diverse student populations. However, concerns are raised about potential biases in the design features of AI-based applications, which may neglect or misrepresent the specific characteristics and learning needs of neuro-divergent students. This study presents the perceptions of neuro-divergent and neuro-typical students (*N* = 18) regarding the inclusivity of the AI-based application *Honest Mirror*. This prototype was designed to enhance students’ presentation skills by providing AI-based feedback. A qualitative approach was adopted, and data were collected using semi-structured interviews. Data were analyzed according to the three design principles of the framework for inclusive AI learning design: engagement, representation, and action and expression. First, concerning engagement, results suggest that mainly neuro-divergent students emphasize the lack of human warmth in AI-based feedback. Second, regarding representation, neuro-divergent students seem to prefer visually minimal, calming color schemes that help sustain their attention. Further, both groups of students indicate that the explainability of the AI system and automated feedback remains unclear. Third, regarding action and expression, both groups of students emphasize that the application is easy to use intuitively. However, suggestions for improvements concern adding explanations of the different steps and actions throughout the application. From a scientific perspective, the findings of this empirical study indicate that inclusivity of AI-based applications might be promoted by incorporating a sense of human warmth into the feedback process. Additionally, understanding the perceptions of neuro-divergent and neuro-typical students regarding different aspects of inclusivity might help to refine theoretical design features of AI-based applications. Practically, this research provides valuable preliminary insights for creating AI-based applications that aim to promote inclusivity among diverse student populations.
