AbnormalAIze – Online Journey to Challenge Stereotypes in AI-Generated Images
Laura CALOGIOIA, Gyan DOOKIE, Katerina GIOTOPOULOU, Theodora MALLIAROU, Elsie RUSSIER
Concept
This workshop helps participants critically examine how artificial intelligence reproduces stereotypes in visual content and learn how to create inclusive and ethical AI-generated imagery. Through creative, collaborative, and gamified activities, students analyze, produce, and reflect on AI images, ultimately designing awareness-raising posters and guidelines for ethical AI use.
Priority
Priority 2: Addressing AI-Generated Biases and Ethical/Deontological Implications Priority 3: Developing Skills as a Critical AI Prompter (in Media Reception and Production)
Objectives
To strengthen critical and creative thinking skills in analyzing and producing AI-generated images.
To critically analyze how AI tools reflect and reproduce social biases and inequalities in visual media.
To experiment with AI prompting as a creative method for questioning and transforming stereotypical representations.
To create diverse, inclusive, and ethically grounded AI-generated images that promote fairness and representation.
Target group
High school students and teachers (adaptable for university and youth groups)
Modalities: steps and instructions
Detailed Explanation of the Modalities (Steps and Instructions)
1. Introduction & Icebreaker (20 min)
Goal: Activate curiosity and introduce AI bias.
Activity: Show a short opening video about stereotypes in AI-generated images.
Game: “Human or AI?” – students are shown 10 images and must guess which were made by humans and which by AI.
Discussion: What made them think one was AI? Were there patterns (e.g. gender, ethnicity, professions)?
Outcome: Awareness that AI reproduces social biases.
2. Identify Stereotypes (25 min)
Goal: Recognize how stereotypes appear visually.
Role Play: Teachers use AI tools to generate images, while students act as “critics.”
Challenge: Each group gets 3–4 AI-generated images and must identify visible stereotypes (gender, age, race, culture, profession, etc.).
Output: Teams create a short post-it wall of “Stereotype Types” (e.g. “Women as nurses, men as scientists”).
Reflection: How do prompts influence AI bias?liste adapté, a notre charte graphique
3. Identify AI Stereotypes (25 min)
Goal: Understand how AI perceives and describes the world.
Challenge: “Human or AI creation?” – analyze both human and AI-generated pictures.
AI interaction: Students ask AI to describe these images; they analyze the generated text and discuss whether the AI reinforced stereotypes.
Challenge: “Human or AI creation?” – analyze both human and AI-generated pictures.
4. Using AI (30 min)
Goal: Practice ethical and creative prompting.
Activity 1: Newsroom Simulation. Students act as journalists creating a visual for a news story. They must generate an AI image that illustrates the topic.
Activity 2: Ethical Prompting Lab. Students identify biases in their image and rewrite the prompt to create a fairer, more balanced version.
Output: Two side-by-side AI images – biased vs ethical.
5. Analyze Stereotypes (30–40 min)
Goal: Deepen analytical skills through a game.
Game: “Dixit – Spot the Bias.” Inspired by the card game Dixit. Groups receive a deck of AI-generated images and a set of prompts. – Match each image with the most likely prompt. – Discuss what stereotypes or assumptions guided the match. – Points for original insights or unexpected connections.
Outcome: Understanding that prompts shape representation.
6. Reflection on Use (20 min)
Goal: Think about responsibility and ethics.
Discussion: What did we learn about AI’s process of image generation?
Production: Co-create a short “AI Guidelines Charter” for their school – e.g., “Use inclusive prompts,” “Avoid gender bias,” “Be transparent when using AI images.”
Output: Shared document (online pad or printed poster)
7. Final Production (30 min)
Goal: Apply learning in a creative way.
Task: Create a digital or printed awareness-raising poster using AI images that challenge stereotypes.
Optional: Display the works in a “Reimagining AI” gallery or online forum.
8. Evaluation (20 min)
Goal: Assess learning through peer feedback.
Peer-to-Peer Evaluation: Students review posters in the online forum, leaving constructive comments based on creativity, inclusivity, and ethical AI use.
Gamified Feedback: Award badges like “Bias Buster,” “Creative Prompter,” and “Ethical AI Guide.”
Gamification Summary
Points: For quick bias spotting, insightful analysis, and creative prompts.
Badges: Earned through specific achievements.
Leaderboard: Optional and friendly – used to boost engagement.
Final Outputs
Alternative AI Image Gallery – stereotypes reimagined
School AI Guidelines – co-created by students
Printed / Digital Exhibition – showcasing inclusive visuals
Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Education and Culture Executive Agency (EACEA). Neither the European Union nor EACEA can be held responsible for them.
