Level 10 · AI Safety & Society · Capstone

Use AI Wisely.
Stay Safe.

Deepfakes. Misinformation. Voice fraud. Privacy. Learn to spot AI misuse, protect your family, and become a responsible AI user — the final step in your AI journey.

⏱ 2–3 hours 🛡️ 5 Safety Skills 🇮🇳 India-specific examples ₹500

🎉 Early Access — ~~₹500~~ Free for first learners

Five learners gathered together, empowered and united around AI safety

scroll to begin ↓

Returning Cast

Five learners. Five real AI safety scenarios.

You have followed them through 9 levels. Now they face the final challenge — using AI responsibly in a world where not everyone does.

🎓 Kavya — Warangal

Gets a shocking WhatsApp video of a politician saying something controversial. It looks real. It is a deepfake. She almost forwarded it to 50 contacts.

💼 Srinivas — Hyderabad

A client asks him to use AI-generated "research" with made-up statistics in a marketing article. He has to decide what is the right thing to do.

🍳 Padma — Vijayawada

Gets a voice call from someone who sounds exactly like her daughter — asking for ₹5,000 urgently. It is an AI voice-clone scam.

📸 Raju — Nizamabad

Accidentally creates an AI photo of a real person without consent for a client poster. He learns about privacy rights and how to fix the mistake.

📚 Meera — Hyderabad (New)

19-year-old college student. Uses AI to write an essay and submits it as her own. Faces a difficult conversation with her teacher — and learns from it.

Act 1

The community meeting: "AI almost got me"

All five characters meet at a community space in Hyderabad. The topic: "AI Bad Experiences." Each person shares something that happened to them or someone they know in the last month. The room goes quiet as the stories come out — one by one.

Kavya: "I almost shared a deepfake video. My cousin shares things without checking. I nearly did the same."

Srinivas: "A client asked me to use AI statistics I could not verify. I almost did it just to keep the client happy."

Padma: "I got a call from someone who sounded exactly like my daughter. I nearly transferred ₹5,000."

Raju and Meera share too. Then the group agrees on something important: the risks of AI are real, but they are learnable. Knowing what to look for is the first and most powerful defence.

Key message: "Knowing the risks makes you stronger, not scared. Every person in this room is already safer than they were before today."

🛡️ Why this level matters

AI gives people powerful tools. But powerful tools can be misused — by others, and sometimes by us without realising it. This level teaches you to protect yourself, your family, and your community.

Act 2

Kavya's deepfake: "It looked completely real"

Kavya looking sceptically at a suspicious video on her phone

🎓 Kavya · Warangal

The video showed a local politician making a shocking statement. The audio was clear. The face looked right. Kavya's thumb was already on the forward button. Then she noticed something: the lips did not quite match the words at one point. She paused. She looked more carefully.

Warning signs she spotted: slightly blurred edges around the face, one moment where the lip movement did not match the audio, skin texture that looked too smooth in bright light.

🎓 Kavya checks before sharing

Instead of forwarding, she went to AltNews.in — India's fact-checking platform. The video was already flagged as fake. A real news outlet had already debunked it. She shared the fact-check link instead, with a message to her family group.

Result: 50 contacts did not receive misinformation. Kavya's family group now has a new rule: check before you share.

How to spot a deepfake video

1Watch the lips: Do they match the words exactly? Deepfakes often have slight audio-visual mismatch.
2Check the edges: Look at the hairline, ears, and chin. Deepfakes often have slight blurring or warping.
3Watch the eyes: Unnatural blinking or staring? AI-generated faces blink less naturally.
4Search before sharing: Paste the claim in Google or AltNews.in. If it is viral, someone has already checked it.
5Pause the "shock reflex": If a video makes you feel outraged or shocked, that feeling is exactly when you must slow down.

💡 Skill 1: Spot a deepfake and verify before sharing

Fact-check tools for India: AltNews.in, BoomLive.in, FactCheck.in, and Google's Fact Check Explorer. Bookmark at least one. Use it every time a video surprises you.

Act 3

Srinivas and the AI-written statistic that was not real

Srinivas cross-referencing an AI article with fact-check websites

💼 Srinivas · Hyderabad

A client asked Srinivas to write a marketing article about health benefits of their product. He used AI to research and draft it. The AI produced this confidently: "85% of Indian families prefer natural health remedies over prescription medicine." It sounded perfect. It was completely made up.

AI hallucination: AI generated a convincing, specific statistic with no real source. This happens regularly. The danger is that it sounds exactly like a real fact.

💼 Srinivas catches it

He searched for the statistic in three places: Google, the Ministry of Health website, and a journal. Nothing. He went back to AI and asked: "What is the source for this statistic?" The AI admitted it could not find one. He removed it, found a real verified statistic, and rewrote the section. He told the client.

Client's response: "Thank you for catching that. I trust your work more because you checked."

❌ If he had not checked

Published a false statistic. Client's reputation damaged. Potential legal issues. Loss of trust.

✅ Because he checked

Article accurate. Client trust increased. Srinivas's professional reputation protected.

💡 Skill 2: Fact-check AI output before using it

Golden rule: AI can write confidently and be completely wrong. For any important claim, number, or statistic from AI — verify in at least one other reliable source. Use Google, government websites, or established news outlets. Never publish AI statistics without checking.

Act 4

Padma's fake daughter call — and Raju's consent mistake

🍳 Padma · Vijayawada

Padma got a call. The voice was her daughter's — the exact same tone, accent, and way of speaking. "Amma, I have an emergency, please transfer ₹5,000 right now." Padma's hand moved to open her UPI app. Then she paused. Something felt wrong.

How it works: Criminals collect voice clips from social media reels and WhatsApp status videos. AI tools can clone a voice from just 30 seconds of audio. The call was AI-generated.

Padma on a suspicious phone call while her daughter is safe in the background

🍳 Padma's safe response

Padma hung up. She called her daughter directly on her known number. Her daughter was home, completely fine, unaware of any call. The scam was confirmed. Padma then told her family about a new rule they set together.

Family safety code: If anyone calls asking for money urgently, they must use the family code word ("Gongura"). If they do not know it, the call is suspicious. Hang up and call back on the known number.

📸 Raju · Nizamabad

A client asked Raju to make a promotional poster for their new hotel. Raju used AI to put a smiling young woman's face on the poster — a face from a reference photo the client gave him. The young woman had not given consent. When she saw the poster displayed publicly, she was upset and asked for it to be removed.

Raju realising his consent mistake with a client poster

📸 Raju learns and fixes it

Raju removed the poster immediately. He apologised. He replaced it with an AI-generated fictional person instead — no real face, no identity. He now has a one-line rule for every client: "We only use real faces with written permission."

India's IT Act protects personal data including photographs. Using someone's face without consent — even for commercial use — can lead to legal complaints. Always ask. Always get consent first.

💡 Skill 3: Protect personal privacy — yours and others'

For voice fraud: Limit personal voice clips in public reels. Set a family code word. Always call back on a known number before any transfer.

For consent: Never use someone's face, voice, or personal images with AI without their explicit permission. What AI can do and what you are allowed to do are not the same thing.

Act 5 · The Capstone

Meera's lesson — and the Responsible AI Pledge

Meera in a mentoring conversation with her teacher about AI and academic integrity

📚 Meera · Hyderabad

Meera used AI to write a full essay for her sociology class. She submitted it as her own work. Her teacher noticed the writing style was different from Meera's usual work. Instead of punishing her, the teacher had a private conversation.

Teacher: "Meera, I'm not angry. AI is a powerful tool. But submitting someone else's thinking — even AI thinking — as your own does not help you grow. What did you actually think about this topic?"

📚 What Meera learned

Meera rewrote the essay — this time, she used AI as a brainstorming partner, not a ghostwriter. She wrote her own analysis. She cited where AI helped her think. Her teacher gave her full marks for the rewrite. More importantly, she understood the difference between using AI and relying on AI.

Right way to use AI for learning: "AI, help me think through the arguments for this topic." Then write your own conclusion. Declare that you used AI as a thinking aid.

The 5 rules of responsible AI use — illustrated pledge card

🤝 The Responsible AI Pledge

🔍 Rule 1: Verify before you share. A shocking video or message deserves 60 seconds of checking before it reaches 50 contacts.

💬 Rule 2: Declare when AI helped you. If AI wrote it, helped plan it, or generated it — say so. Honesty builds trust.

❌ Rule 3: Never use AI to harm or deceive. Deepfakes, fake reviews, manipulative content — if it could hurt someone, do not create it.

🔒 Rule 4: Protect personal data. Yours and others'. Consent before using faces, voices, or private information with AI tools.

👥 Rule 5: Teach one person. You now know more than most people around you. Share one thing you learned today with someone who needs it.

Your Safety Toolkit

5 skills every responsible AI user has

Save this. Share it. Use it every time AI surprises you.

🎭 Spot Deepfakes

Check lips, edges, and eyes. Pause the shock reflex. Search before sharing. Use AltNews.in or BoomLive.in.

📊 Fact-Check AI Output

Never publish AI statistics without verifying. Ask AI for its source — it will often admit it has none.

📞 Voice Fraud Defence

Set a family code word. Always call back on a known number. Never transfer money based on an unexpected call.

🖼️ Consent Before AI Images

Never use real faces with AI without permission. India's IT Act protects personal images. Always ask first.

🎓 Honest AI Use

Use AI as a thinking partner, not a ghostwriter. Declare when AI helped. Your thinking is still what matters most.

Practice

Three things to do today

Before you take the quiz, try at least one of these.

Challenge 1 — Set your family code word

Talk to your family today. Agree on a secret code word that anyone calling about money must know. Write it somewhere safe. Test it once with one family member.

Challenge 2 — Fact-check something

Go to AltNews.in or BoomLive.in. Search for any viral video or claim from the last week in your WhatsApp groups. See if it is verified, debunked, or unverified. Share what you find with one person.

Challenge 3 — Teach one rule

Share one of the 5 Responsible AI Rules with one person today — a parent, sibling, friend, or colleague. Explain it in simple words. You become a teacher the moment you do.

🏆

Level 10 Complete!

Responsible AI User

You have completed all 10 levels of Mitra AI Life Education. You can spot deepfakes, fact-check AI output, protect your privacy, use AI with integrity, and teach others around you. That is the full journey.

🎭 Deepfake Detection 📊 Fact-Checking 📞 Voice Fraud Defence 🖼️ Consent & Privacy 🎓 Honest AI Use

🌟

You completed the full journey!

From your first AI prompt in Level 1 to becoming a responsible AI advocate in Level 10 — you have come an incredible distance. India needs people like you.

← Back to Home

Use AI Wisely.
Stay Safe.

What you will learn in Level 10