AI Ethics, Fairness & Explainability (Level 3)
Ethical issues that arise when AI systems are developed and deployed.
Key ethical issues​
- Bias and fairness: training data may reflect historical inequalities leading to unfair outputs.
- Privacy: collecting and using personal data can harm individuals if misused.
- Transparency: many models are opaque, making it hard to explain decisions.
- Accountability: who is responsible when an automated system causes harm?
Explainability​
Explainability is about making models and their decisions understandable. Approaches range from simple, interpretable models to tools that explain complex models' behaviour.
Classroom exercise​
- Find a short news article where AI caused controversy (bias, privacy, or accountability). Write 5 bullets summarising the issue and suggest one technical or policy mitigation.