Workshop explaining how AI can reproduce discrimination through biased data, model design, and existing social structures. It treats discrimination as both technical and political, and provides materials for educators to help participants detect, understand, and counter algorithmic bias.
Why it matters for ENACT-AI
The resource encourages young people to understand how AI systems learn from data, how human biases can become embedded in algorithmic systems, and why AI technologies are not necessarily neutral or objective. It also connects AI literacy with ethics, social justice, power structures, intersectionality, and democratic debate about how technologies should be designed and governed.
Learning outcomes
Learners will be able to explain how algorithmic systems and machine learning can reproduce existing social biases, identify examples of discrimination through AI, understand the relationship between data and algorithmic decision-making, recognise the importance of intersectionality when analysing discrimination, critically evaluate whether technologies can be considered neutral, identify ethical questions related to AI-based decisions, and reflect on how AI should be designed and governed in a fair and responsible way.
Suggested activity
Divide participants into small groups and assign each group a different application area, such as recruitment, healthcare, facial recognition, or education. Ask them to investigate how machine learning is used in that area, what data the system relies on, who may benefit or be disadvantaged, and what forms of discrimination could occur. Each group presents its findings and proposes principles or safeguards that should guide the development and use of AI systems.