This study examines the impact of different avatar pictures (gender & disability representation) and gendering on students’ perceptions of chatbots in an interaction on learning strategies with 180 students from a German university. In the first experiment, we manipulated the chatbot’s humanoid profile picture based on gender and the representation of a visible handicap (wheelchair). In the second experiment, we varied its language style. Statistical analysis revealed that displaying a physical disability significantly enhanced trust, credibility, and empathy but reduced perceived competence and dominance. Gender-sensitive language improved perceptions of competence, trust, credibility, and empathy, whereas we did not find significant interaction effects between both factors. Our results imply the necessity of a more inclusive design of information systems and highlight designers’ responsibility in raising awareness and mitigating unconscious bias, as digital learning (technologies) continue to advance.