Class 7
Chapter 4: Ethics and AI Bias Awareness
AI learns patterns from data. If the data only represents one group of people, the AI will work poorly for everyone else. This is called bias, and avoiding it is part of using AI responsibly.
Learning outcomes
- Explain how biased data leads to biased AI
- Suggest steps to make AI use responsible
Activities
- Spot-the-bias scenario discussion
Worked examples
Read through these first, then try the practice below.
Example 1 — Spot the bias
A face-recognition AI is trained only on photos of adults. Why might it work poorly on children?
Solution: Children's faces look different from adults' and were never shown to the model, so it cannot recognise them well.
Example 2 — Fix the bias
How could we improve the model in Example 1?
Solution: Add many photos of children of different ages, lighting and backgrounds to the training data, then retrain.
Self-do practice
Question 1 of 3 · Score 0/0True or False: An AI is fair if its training data represents all the people who will use it.
Sign in as a student in this class to save your progress.