📖 Chapter 6 of 9 · AI For New Learner Advance 1 (Multi Indian Language) · Class 7
Chapter 6: Bias, Truth & Deepfakes
In this chapter
- Veer from Dehradun catches a fake video.
- Anika from Vizag fights BIAS in AI.
- What is a deepfake? How to spot one.
- Why fairness matters — AI can be UNFAIR if data is unfair.
- Your duty as a smart AI citizen.
The dark side of AI
AI is powerful, but it can also be MISUSED. Deepfakes (fake videos that look real) can fool millions. Biased AI can be unfair to certain people. This chapter is serious — but we MUST learn it. The smartest AI users are not the ones who use it most, but the ones who SPOT problems and STOP harm.
Did you know?
In 2024, a deepfake video of a famous Indian politician spread fake news to 1 crore phones in 6 hours, just before elections. The Election Commission asked WhatsApp to limit forwards. Lesson: fake videos can change the country. Spot them, don’t share them.
Story 1: Veer Catches a Deepfake!
From: Dehradun, Uttarakhand · Veer, age 12
Veer lives in Dehradun, in the cool foothills of the Himalayas. One night his Papa received a WhatsApp forward — a video of a famous Bollywood actor saying: “Send Rs 500 to this number — you will get back Rs 5000! Special offer!”
Papa was tempted. But Veer noticed strange things. The actor’s left eye was bigger than the right. The mouth moved but the chin didn’t match. The lighting on the face was different from the background. “Papa, this is FAKE. A deepfake.”
Papa was shocked. “You can tell?” Veer showed him: “My ICT teacher told us — check the eyes, the mouth, the lighting, the voice. If even ONE thing is weird, it’s probably a deepfake.” Veer used a website to upload the video — it confirmed: 99% deepfake.
Papa did not send money. Papa did not forward the video. Instead, he sent a message in the family group: “Be careful — this is a deepfake.” Veer felt proud. He had saved his Papa’s money AND helped his whole family stay safe.
Story 2: Anika the Bias Hunter
From: Visakhapatnam, AP · Anika, age 12
Anika lives in Vizag (Visakhapatnam) — port city of Andhra. She loves science. Her favourite uncle is a doctor — and her favourite aunt is also a doctor. So when her class did an AI art project — “AI, draw 5 doctors” — Anika was shocked. The AI drew 5 MEN, all looking similar. Not one woman. Not one Indian-looking face.
Anika went home angry. “Why is AI like this?” Her father, a software engineer, explained: “AI learns from PICTURES it was trained on. If 90% of doctor photos on the internet are white men, then AI thinks ‘doctor = white man’. This is BIAS.”
Anika did a project. She showed: AI Bias against women in medical pictures. AI Bias against dark skin in beauty filters. AI Bias against Indian names in resume screening. The class was shocked!
Her teacher pinned the project on the school board. Anika said: “We must give AI FAIR data — pictures of women doctors, dark-skinned models, Indian names. Then AI will be fair to ALL. Building unbiased AI is a NEW kind of national service.”
Solved illustrations
Truth, fairness, and AI ethics
1. What is a deepfake?
Show answer
Answer: A video, audio, or image generated by AI that looks REAL but is FAKE. AI swaps faces, copies voices, or invents people. Deepfakes can be funny (movie effects) or dangerous (fake politician videos, fake celebrity scams).
2. How can you spot a deepfake?
Show answer
Answer: Check: (1) Eyes — do they blink naturally? Are they the same size? (2) Mouth — does it match the words? (3) Lighting — does the face match the background light? (4) Voice — does it sound robotic? (5) Source — is it from a trusted news channel? If unsure, DON’T SHARE.
3. What is ‘bias’ in AI?
Show answer
Answer: When AI is UNFAIR to certain groups — because its training data was unfair. Example: if AI is only shown male doctors, it thinks doctors are male. This is harmful — it spreads stereotypes and stops fair opportunities.
4. How do we FIX biased AI?
Show answer
Answer: Give AI FAIR and MIXED data: women doctors AND men doctors. Dark skin AND light skin. Indian names AND foreign names. Old AND young. Rural AND urban. Fair data → Fair AI.
5. What should you do if a video looks suspicious?
Show answer
Answer: (1) Don’t forward. (2) Check the source. (3) Look up on official news sites. (4) Tell a parent or teacher. (5) Report to WhatsApp/Instagram. Remember: forwarding a fake video makes YOU part of the problem.
Try this at home
Pick ONE dilemma to discuss: Should AI be allowed to grade exams? Should deepfakes ever be legal? Take 10 minutes to write your view. Then ask a parent and a friend. Compare 3 answers — see how different people think differently.
Questions — Chapter 6
Citizenship questions — for thinkers, not just users
A. Fill in the blanks
- A fake video made by AI is called a .
- When AI is unfair to certain people, we call it .
- Veer’s home city is , in the state of .
- To fix biased AI, we need fair and data.
- If you see a suspicious video, you should forward it.
B. Choose the BEST answer
- The first thing to do with a strange WhatsApp forward is — (a) share with all friends (b) click any link inside (c) check the source and STOP
- If an AI keeps showing only male engineers, it is — (a) just lucky (b) biased (c) very smart
- The BEST way to fight deepfakes is — (a) believe nothing (b) think before sharing, check sources (c) delete WhatsApp
C. Long Answer
- Explain in 4-5 sentences: what is AI bias and how can we fix it?
- Imagine you see a deepfake of your favourite cricketer. What do you do, step by step?
Brain Buster
Be a ‘Truth Watchdog’ for your family. Anytime you see a suspicious forward, check before sharing!
Check your understanding — Chapter 6
Tap an answer to check it instantly. These are for practice — nothing is saved or sent.
Q1. What is a deepfake?
- A deep swimming pool
- An AI-made video, audio or image that looks real but is fake
- A type of camera
Show answer
B. An AI-made video, audio or image that looks real but is fake
Q2. What is 'bias' in AI?
- AI being very fast
- AI being unfair to certain groups because its training data was unfair
- AI running out of battery
Show answer
B. AI being unfair to certain groups because its training data was unfair
Q3. How do we FIX biased AI?
- Turn it off
- Give it fair, mixed data — all genders, skin tones, ages, regions
- Use it less
Show answer
B. Give it fair, mixed data — all genders, skin tones, ages, regions
Q4. The first thing to do with a strange WhatsApp forward is —
- share with all friends
- click any link inside
- check the source and STOP
Show answer
C. check the source and STOP