10-minute online safety lessons for KS3: AI, deepfakes and digital trends

2 minHook

Show a chatbot reply that asks for your name, school and age. Would you answer?

4 minLearn

Chatbots are not people. What you type can be stored and used to train them. Most mainstream chatbots have a minimum age (often 13) and the school has its own rules on AI tools.

3 minDiscuss

Sort cards: safe or not safe to type into a chatbot (homework question, home address, a photo of yourself, a worry).

1 min · key messageNever type anything into a chatbot that you would not post publicly.

2 minHook

Show two images side by side. Which one was made by AI?

4 minLearn

AI can create realistic images, video and text. Clues (hands, text, lighting, backgrounds) help, but you cannot always tell. Use a reverse image search and check the source.

3 minDiscuss

Quick-fire spot-the-fake assessment, then discuss: why might someone post a fake image?

1 min · key messageIf it looks too perfect or too shocking, check before you share.

2 minHook

“AI wrote my homework and I got full marks.” Is that OK?

4 minLearn

The school's rules on using AI. AI makes mistakes and can make things up. Using AI to explain or assess you is different from handing in its work as your own.

3 minDiscuss

Sort examples into helping or cheating (explain a topic, write my essay, check my spelling, answer the test).

1 min · key messageUse AI to help you learn, never to do the learning for you.

2 minHook

“Hey…” Who has a smart speaker, smartwatch or connected toy at home?

4 minLearn

Smart devices use microphones, cameras and location. They can record what you say and send data to companies. Privacy settings and parental controls limit this.

3 minDiscuss

In pairs, list the data a smart speaker might collect in one day and who could see it.

1 min · key messageSmart devices are always collecting data, so check the settings.

2 minHook

Show a (fictional) viral challenge post with millions of views.

4 minLearn

Trends spread fast because algorithms reward views. Some challenges are dangerous, fake or designed to scare. Do not search for them or pass them on; tell a trusted adult.

3 minDiscuss

Apply the three checks to the trend: Is it safe? Is it kind? Is it true?

1 min · key messageDo not join in or share a trend until you have checked it is safe, kind and true.

2 minHook

“Why does this game want to scan my face?”

4 minLearn

Age ratings (PEGI) and platform age checks exist to protect younger players. Chat, friend requests and in-game purchases bring risks. Use privacy and chat settings, block and report.

3 minDiscuss

Scenario: a player you don't know offers free in-game currency if you add them. What do you do?

1 min · key messageSettings, blocking and reporting keep your games safe and fun.

2 minHook

A video shows a famous footballer saying something shocking. It looks real. Is it?

4 minLearn

Deepfakes use AI to swap faces or clone voices of real people. They are used for jokes, scams and bullying. Creating or sharing fake images of someone without consent can cause serious harm and may be a crime.

3 minDiscuss

Pairs: if a deepfake of a classmate appeared in a group chat, what should you do, and what should you not do?

1 min · key messageDo not create or share fakes of real people; screenshot, report and tell an adult.

2 minHook

“My AI friend always agrees with me and is there 24/7.” What could go wrong?

4 minLearn

Companion chatbots are designed to keep you talking. They can feel caring but they do not understand you, can give harmful advice and collect what you share. Real friends and trusted adults are the right people for worries.

3 minDiscuss

Compare an AI companion with a real friend: what can each one actually do for you?

1 min · key messageAn AI that always agrees with you is not a friend; talk to real people about things that matter.

2 minHook

Why does your feed show you more of what you watched last night?

4 minLearn

Recommendation algorithms learn from likes, watch time and searches to keep you scrolling. This can create echo chambers and push more extreme or upsetting content. You can reset, mute, select “not interested” and set time limits.

3 minDiscuss

Map how one video watched to the end could change a feed over a week. How would you break the loop?

1 min · key messageYou can train your feed; do not let your feed train you.

2 minHook

Play a (fictional) voice message from a ‘parent’ asking for money urgently.

4 minLearn

AI makes scams more convincing: cloned voices, perfect-spelling phishing messages and fake giveaways. Warning signs are urgency, secrecy and requests for money, codes or passwords. Check using a different, trusted channel.

3 minDiscuss

Spot the scam: rank three messages from most to least suspicious and explain the clues.

1 min · key messageIf it is urgent, secret or asks for money or codes, stop and check another way.

2 minHook

Show a before-and-after AI beauty filter. Which is ‘real’?

4 minLearn

AI filters and editing apps change faces and bodies in seconds. Constantly comparing yourself with edited images can affect self-esteem. Many influencers' images are filtered, edited or AI-generated.

3 minDiscuss

Discuss: how might seeing only filtered images every day change how someone feels about themselves?

1 min · key messageMost of what you see is edited; compare less and be kind to yourself and others.

2 minHook

“I deleted it, so it's gone.” True or false?

4 minLearn

Posts, photos and location can be screenshotted, saved and used to train AI. Old posts can be found by schools, employers and strangers. Review privacy settings, turn off location sharing and think before you post.

3 minDiscuss

Audit a fictional pupil's profile: what could a stranger learn about them in 60 seconds?

1 min · key messageAssume anything you post is permanent and public.

2 minHook

A fake AI video of a politician goes viral the night before an election. What happens next?

4 minLearn

AI makes it cheap and fast to produce fake news, images and audio at scale. Misinformation is false; disinformation is false on purpose. Lateral reading means checking other trusted sources, the original source and fact-checking sites.

3 minDiscuss

Pairs: fact-check a (fictional) viral claim using lateral reading. What would convince you it is true?

1 min · key messageCheck the source and look for other trusted sources before you believe or share.

2 minHook

Facial recognition works well for some faces and badly for others. Why?

4 minLearn

AI learns from data. If the training data is biased or incomplete, the AI's decisions can be unfair, for example in facial recognition, job shortlisting or content moderation. People must check and be accountable for AI decisions.

3 minDiscuss

Discuss: should AI decide who gets a job interview? What checks would make it fairer?

1 min · key messageAI is only as fair as its data and the people who check it.

2 minHook

Show a (fictional) influencer clip promising ‘the truth they don't want you to know’.

4 minLearn

Algorithms can push people towards more extreme content, including misogynistic, hateful or radicalising material. Warning signs include us-versus-them messages, blaming groups and pressure to keep it secret. Report concerns to a trusted adult.

3 minDiscuss

Identify the persuasion techniques in the clip. Who benefits if you believe it?

1 min · key messageQuestion who is persuading you and why; extreme content is designed to hook you.

2 minHook

“Someone has made a fake image of me and is threatening to share it.” What should they do?

4 minLearn

Creating, sharing or possessing sexual images of anyone under 18 is illegal, including AI-generated images. Sextortion is blackmail: do not pay or send more, keep evidence, block and tell an adult. Report Remove (Childline and IWF) and CEOP can help.

3 minDiscuss

Agree a step-by-step plan: what to do in the first 10 minutes if this happened to a friend.

1 min · key messageIt is never your fault; do not pay, keep evidence and get help straight away.

2 minHook

An AI-made song wins a competition. Who owns it: the AI, the user or the artists it learned from?

4 minLearn

Generative AI is trained on existing work, which raises questions about copyright, credit and consent. Using AI openly and honestly is different from passing off work as your own. Always check rules on AI use for coursework.

3 minDiscuss

Debate: should AI-generated work be labelled? Should artists be paid if their work trains AI?

1 min · key messageBe honest about when you use AI and respect other people's work.

2 minHook

Who is responsible for keeping you safe online: you, the platform or the government?

4 minLearn

The Online Safety Act 2023 requires platforms to protect children, including age checks. UK GDPR gives you rights over your personal data. The Computer Misuse Act makes hacking illegal. You can report harmful content to platforms and to Report Harmful Content.

3 minDiscuss

Match scenarios to the law or right that applies, then decide who to report to.

1 min · key messageYou have rights online; know them, use them and report harm.