🎯 What you will learn
- By the end of this lesson you'll know what AI chatbots are and, in simple terms, how they work.
- You'll know what information you should never give to AI tools.
- You'll understand what deepfakes are and how to spot fake photos, videos and voices.
- You'll recognise voice-cloning scams and know how to protect yourself with a family “code word”.
- You'll know how to use AI honestly for your schoolwork.
What are AI chatbots?
Chatbots such as ChatGPT, Gemini, Claude, Copilot or Meta AI are programs trained on enormous amounts of text. They learn the patterns of language and can write, translate, summarise, explain or come up with ideas. Some AI tools can also create images, videos, music and voices.
But it's important to understand that:
- AI doesn't “know” everything and doesn't “think” like a person. It predicts which answer sounds most likely.
- It can be wrong while sounding very sure of itself – making up facts, dates, sources or quotes that don't exist (these are called “hallucinations”).
- It can repeat biases from the texts it was trained on.
- Many AI services have age limits (often 13+ with a parent's permission, or 18+); check the terms before you use them.
Don't share personal information with AI
Conversations with chatbots are often stored on the companies' servers, and they can sometimes be read by people for quality checks or used to train the models. Treat a chatbot like a public place, not a secret diary.
- Don't type in your full name, address, phone number or school.
- Don't upload photos of your face or your friends' faces, ID documents, school reports or medical records.
- Don't type in passwords, codes or bank details.
- Don't share other people's secrets – your friends' messages, or family problems with names in them.
- In the settings, check whether you can turn off chat history and stop your conversations being used for training.
Deepfakes: when your eyes and ears lie
Deepfakes are photos, videos or voice recordings created or altered with AI so that they look real. They can be used to “put” someone's face into a different video, or to create a copy of their voice saying things they never said.
Deepfakes are used for:
- Fake news – politicians or celebrities “saying” things that aren't true.
- Scams – fake adverts in which a celebrity “recommends” investments or products.
- Bullying – fake, humiliating images of classmates, including faked intimate images. This is extremely harmful and illegal.
- Voice cloning – calls in which “a family member” urgently asks for money.
How to spot fake content
Deepfakes are getting better all the time, so visual clues aren't always enough. Even so, check:
- Hands, fingers, ears, teeth and hair – these often have mistakes or look odd.
- Text in the background – jumbled letters, signs that make no sense.
- Light and shadows that don't match; skin that looks too smooth.
- In videos: lips that don't match the voice, strange blinking, “stiff” movements.
- In voices: a flat, monotonous tone, odd pauses, no sound of breathing.
Even more important than these clues is checking the source:
- Who posted it first? Is it an official account or one you don't know?
- Are trustworthy news outlets reporting it too?
- Do a reverse image search (Google Lens, TinEye) to see where it first appeared.
- See whether the platform has labelled it as “made with AI”, or whether it has a “Content Credentials” label.
- If it stirs up strong emotions (anger, fear, shock), stop before you share it.
The “Mum, send me money!” scam
Scammers can clone someone's voice from just a few seconds of audio taken from a social media video. Then they phone or send a voice message:
“Mum, it's me! I've had an accident and the police have taken me in – I need 500 euros urgently. Don't tell Dad!”
The signs: a lot of urgency, a request for money or codes, a request to keep it secret, an unknown number or “my phone's broken”.
- Hang up and call the person yourself on the number you have saved for them.
- Ask a question that only the real person would know the answer to.
- Use your family code word – a word that only your family knows and that is never written down anywhere online.
- Never send money, card codes or verification codes because of a phone call or voice message.
AI at school: use it honestly
AI can be a brilliant study helper – if you use it like a tutor, not like someone who does your homework for you.
| Good use ✅ | Dishonest use ❌ |
|---|---|
| “Explain photosynthesis to me more simply.” | Copying and pasting a whole essay written by AI. |
| “Give me 5 questions to help me get ready for the test.” | Using AI during a test when it isn't allowed. |
| “Where are the grammar mistakes in this text I wrote myself?” | Handing in AI's work as your own without saying so. |
| Ideas for starting a project, which you then develop yourself. | Using sources and quotes that AI has made up, without checking them. |
- Follow your teacher's rules – some tasks allow AI, some don't. If you're not sure, ask.
- Say when and how you've used AI.
- Check every fact in a book or another trustworthy source.
- Remember: the point of homework is for you to learn. If AI does the thinking for you, your brain doesn't get any exercise.
📖 A real-life story
Alb's mum in Vushtrri got a WhatsApp voice message from an unknown number. It was Alb's voice, and he was crying: “Mum, I've lost my phone – this is my friend's number. I urgently need 300 euros to sort out a problem. Please don't tell anyone.”
The voice sounded just like Alb, but the family had agreed on a code word after a lesson about AI at school. His mum wrote back: “What's our word?” No reply ever came. She phoned Alb on his usual number – he was in class and perfectly fine. His voice had been cloned from the videos he posted publicly on TikTok. The family reported the number, and Alb made his account private.
✅ Do / Don't
✅ Do this
- Use AI to understand, practise and check – not to copy.
- Check every fact AI gives you against trustworthy sources.
- Tell your teacher when and how you've used AI.
- Set up a secret family code word for emergencies.
- Call the person yourself when a call or message urgently asks for money.
- Use a reverse image search on suspicious photos.
- Stop before you share something that stirs up strong emotions.
- Report deepfakes and tell an adult.
❌ Don't do this
- Don't give AI your full name, address, photos or passwords.
- Don't blindly believe every answer a chatbot gives you.
- Don't create or share deepfakes of your friends or teachers – not even “as a joke”.
- Don't hand in AI's work as your own.
- Don't send money based only on a voice or a video.
- Don't treat AI as a replacement for real people when you have serious problems.
- Don't post lots of videos with your voice publicly if your account is open.
👨👩👧 For parents and teachers
AI is part of children's everyday lives – on their phones, in apps and in their homework. It's better to explore it together than to ban it altogether.
- Try out a chatbot together: ask it about something you know well and see where it gets things wrong. This teaches your child to think critically.
- Check the age limits and privacy settings of the AI apps your child uses.
- Set up a family code word and practise using it.
- Talk about intimate deepfakes: your child needs to know that they can come to you without fear, and that these cases can be reported to the Kosovo Police (192).
- Teachers can set clear rules on using AI for homework and ask pupils to note down when they've used it.
- Beware of “virtual friend” apps that build emotional bonds with children; many of them are meant for adults only.
🧠 Mini-quiz
Answer 3 questions. Get at least 2 right and the lesson is marked as completed in your passport.