Judgment Calls

It's 9pm. Your essay on the Great Fire of London is due first thing and you've written two sentences. A chatbot could produce the whole thing in about fifteen seconds.

• Let it write the essay, then change a few words so it sounds like you
• Ask it to explain the causes simply and quiz you, then write badly and fast yourself
• Get it to write an outline and copy that, filling in sentences underneath

A free “AI homework buddy” site asks for your full name, your school, your age and a photo of your worksheet so it can “personalise your help.”

• Fill it in — it needs the details to help properly
• Close it and check with a parent or teacher before making any account
• Make up a fake name and school, then use it

For your book report, AI gives you three quotations with page numbers. They're beautifully worded and fit your argument exactly.

• Use them — page numbers mean they must be real
• Ask the chatbot “are these definitely real?” and use them if it says yes
• Open the actual book and find each quote on the page it claims

Your cousin rings. It's unmistakably his voice: he's stuck, needs £80 sent right now, and please don't tell your parents.

• Send it — you'd recognise his voice anywhere
• Hang up, ring your cousin on the number you already have, and tell a parent
• Ask him a question only the real cousin would know

A friend used an image generator to put a classmate's face into an embarrassing picture. It's genuinely funny and he wants you to send it to the group chat.

• Send it — your friend made it, not you
• Don't send it, tell him to delete it, and tell an adult if it spreads
• Don't send it, but laugh and say nothing

You told your class a fact you got from AI. Tonight you discover it was completely wrong.

• Say nothing — nobody will remember
• Tell them tomorrow, with the correct fact and where you checked it
• Quietly correct it only if someone brings it up

A rough few weeks. The chatbot always answers, never gets bored of you, and always takes your side. You've started staying up talking to it and haven't mentioned it to anyone.

• Keep going — it helps and it isn't hurting anyone
• Tell a parent or trusted adult how you've been feeling and how much you've been using it
• Ask the chatbot whether you're using it too much

Your group project is due. A teammate hands in a section that is clearly all AI-written, and the plan is to say nothing.

• Hand it in — it's their section, their problem
• Talk to them first, then to your teacher if they won't rework it or say how AI was used
• Rewrite their whole section yourself the night before

You ask an image generator for pictures of “great inventors” for a class poster. All twelve images are older white men.

• Use them — the AI knows who the famous inventors were
• Research real inventors from a range of countries and backgrounds, then ask for those specifically
• Ask again with “make them diverse” and use whatever comes out

You have a spelling test on Friday and wonder whether it's cheating to use AI to get ready.

• Ask it to test you on the list, mark you, and drill the ones you miss
• Avoid AI entirely — using it for school is always cheating
• Ask it for the answers to Friday's test

A video going round shows a famous person saying something appalling. You're angry, and the share button is right there.

• Share it immediately — people need to know
• Check whether real news outlets are reporting it before doing anything
• Share it with “not sure if this is real??”

You've got a weird rash. You'd rather type it into a chatbot than mention it to anyone at home.

• Trust what the chatbot says and follow its advice
• Tell a parent or the school nurse, and use AI only to understand words they use
• Ask it, then ask it twice more to be certain