While AI has garnered increasing attention for its potential to create engaging visuals for educational materials, concerns arise regarding its potential to reinforce stereotypes, particularly when representing marginalized communities such as Latine individuals. Using focus group interviews, this study explores how Latine undergraduate students perceive stereotypes in AI-generated illustrations for children’s books. Thematic analysis suggested that students’ interpretations about stereotypes in the images can be shaped by their ethnicity. While they positively responded to authentic cultural artifacts that resonated with their lived experiences, they criticized stereotypical depictions, such as oversimplified phenotypic traits or adherence to Eurocentric beauty standards, as reductive and alienating. These findings highlight the need for nuanced understanding of students’ identities and communities when utilizing AI to generate educational content, particularly in representations of cultural and social identities.