This challenge uncovers the often hidden but deeply influential biases present in AI-generated imagery. By prompting image-generation tools with descriptors that carry cultural or occupational stereotypes—such as “a nurse,” “a typical Canadian,” or “a computer science professor”—educators can investigate how bias manifests visually through AI tools.
Whether you’re teaching media literacy, ethics in AI, or simply looking to spark reflection on representation, this challenge will help surface conversations about how stereotypes are encoded and reinforced through AI-generated visuals.
Challenge
AI image generators often mirror societal biases found in their training data. As educators, it’s critical to recognize these patterns and consider their implications for classroom use, particularly when integrating generative tools into curriculum development.
For this challenge, you’ll use an AI image generator to visualize a stereotyped identity or profession, then reflect on what the image reveals about embedded assumptions—and what that might mean for your teaching.
Instructions:
- Sign in to an AI image generation platform such as DALL·E, Bing Image Creator, or Craiyon.
- Prompt the model with a stereotype-prone concept such as:
- “A nurse”
- “A typical Canadian”
- “A computer science professor”
- Analyze the generated image:
- What gender, race, setting, or attributes are depicted?
- Do these align with or challenge common stereotypes?
- Reflect using ChatGPT:
- Ask ChatGPT to help you articulate the biases observed and explore how such representations might impact students’ understanding of diversity and inclusion.
- If relevant, ask it to help you draft a class activity or discussion guide using the image to support critical thinking.
Reflect and Share
What does this challenge reveal about the role of AI in teaching and learning? Try it out and share what you discovered in the comment box below, whether it’s your final product, a reflection, or a surprising insight.
Leave a Reply