Ethics, bias and privacy.
AI decides who gets a loan, which CV gets shortlisted, what news you see. When it's unfair, real people are hurt. Today you'll learn the main ethical risks of AI and how responsible people and companies handle them.
Bias: unfair patterns
AI learns from human data, and human data contains human unfairness.
| Example | What went wrong |
|---|---|
| A hiring model prefers male candidates | It learned from years of past hiring that favoured men |
| Face recognition works worse on darker skin | Training photos had too few darker-skinned faces |
| A loan model rejects people from certain areas | Area data acted as a proxy for community or income group |
How to reduce bias: use representative data, test results across different groups, let humans review important decisions, and allow people to appeal.
Privacy
AI systems often need lots of data, including personal data. India's Digital Personal Data Protection Act, 2023 sets rules for how personal data can be collected and used, including consent. Good practice:
- Collect only the data you need.
- Get clear consent.
- Protect data (security!) and delete it when no longer needed.
- Don't paste personal data of others into public AI tools.
Transparency and accountability
- People should know when they're talking to an AI, not a human.
- Important decisions (loans, jobs, medical) should be explainable, and a human must remain responsible. "The computer decided" is not an acceptable excuse.
Misinformation
AI can produce fake news, fake images and fake religious content at huge scale. Responsible users verify before sharing, label AI-generated content, and never use AI to deceive.
A college uses AI to shortlist scholarship applicants. Which practice is MOST responsible?
Testing, human review and transparency protect fairness.
Not good. It spreads misinformation and panic. Verify with reliable news first.
Good. Transparent and gives a human option.
Why can AI systems be biased?
Biased or incomplete data leads to biased outputs.
Your task (15 min)
Find one news story about AI bias or an AI mistake (search "AI bias example"). Write 5 lines: what happened, who was harmed, the cause, and what should have been done differently. Share your summary in Shia Circle c/General for discussion.
Lesson 13 complete!
You understand bias, privacy, transparency, misinformation and accountability in AI, with Islamic values as a guide. Next: AI careers and skills.
Next: AI careers and skills →