Who Does AI Think Early Childhood Educators Are? Gender Bias in Image Generation

Abstract Book of the 8th Global Conference on Education and Teaching

Year: 2026

[PDF]

Who Does AI Think Early Childhood Educators Are? Gender Bias in Image Generation

Eloise Thomson, Rachel Chapman, Grant Cooper, Zhara Cooper

ABSTRACT:

Generative artificial intelligence has transformed how we create and consume visual media in education. AI image generation models, particularly DALL-E-3, offer unprecedented opportunities for visual representation in educational contexts. However, these technological advances raise critical questions about social norms and narratives, especially concerning gendered roles and professions in early childhood education.
This study examines patterns in AI-generated images of early childhood educators and leaders through content analysis, focusing on the representation of gender and other characteristics. Drawing on feminist post-structuralist theory, the research investigates whether these tools perpetuate existing stereotypes or provide opportunities to challenge and redefine them.
Findings indicate that generative AI reinforces existing stereotypes associated with gender in professional roles within early childhood education. The analysis reveals that AI-generated images portray educators and leaders in ways that are biased and conform to traditional gender norms, with educators predominantly depicted as female and leaders as male. This research contributes to growing scholarship on GenAI’s impact in educational settings and its role in either reinforcing or dismantling stereotypes. The implications are significant for teacher education, professional identity, and the broader challenge of achieving gender equity in early childhood education leadership.

Keywords: Generative Artificial Intellegience; image analysis; leadership; performativity; stereotypes