Think & create · Lesson 15 of 16
AI, deepfakes & chatbots
Use AI thoughtfully and check claims, voices and images before trusting them.
What is it?
Artificial intelligence (AI) is software that can write text, answer questions or create images, voices and videos. A deepfake is a fake video, image or voice made with AI to look or sound like a real person. A chatbot is an AI you can type or talk with.
Why it matters
AI can be really useful for learning and creating, but it can also be wrong, and it can be used to trick people or harm someone’s reputation. Knowing its limits helps you use it well and spot fakes.
How it works
Chatbots predict likely words based on huge amounts of text, so they can sound confident while being wrong or inventing sources. Deepfake tools learn from real photos and recordings to copy a person’s face or voice. Checking important facts elsewhere and verifying surprising requests through a known contact keeps you safe.
Examples
- Ages 9–11: Oscar asks a chatbot for facts for his project and gets a cool fact about octopuses. He checks it in a library book first, and the book says something different.
- Ages 12–14: A video of a teacher saying something strange spreads around school. Zara notices odd lip movements and that it only appears on one account, so she doesn’t share it.
- Ages 15–17: Luca gets a voice message that sounds like his brother asking for money urgently. He calls his brother directly, and his brother never sent it.
Learn
Check important answers
A chatbot can give a confident wrong answer or invent a citation. Check the original source, especially for facts that affect a decision.
Protect private details
Leave out identifying information, personal photos, passwords and private messages. For personal problems, talk to someone you trust.
Check voices and images
Realistic audio or video can be generated or edited. Verify a surprising request through a known contact method. Do not share someone’s image as a joke or impersonation.
Imagine this
- Ages 9–11: A chatbot gives you a weird fact for a school project.
- Ages 12–14: Someone posts a convincing clip of a teacher saying something surprising.
- Ages 15–17: An AI companion asks for your private messages to give better personal advice.
Practise: what would you do?
1. AI gives you a surprising claim with a link. What should you do?
- Use it because an answer with a link must be reliable
- Open the source and check it really supports the claim
- Ask the chatbot if it is sure and stop there
Show the answer
Open the source and check it really supports the claim. A confident answer or citation can still be wrong.
2. A voice that sounds like a sibling asks you to send money. What is the safer step?
- Send it because you recognise the voice
- Reply using the contact details they give you
- Reach your sibling or a trusted adult through a contact you already know
Show the answer
Reach your sibling or a trusted adult through a contact you already know. A familiar voice is not enough to verify a request.
Videos
- What Are Deepfakes? (ABC Education)