“KI im Alltag” series · Printable
Checklists & Templates
Straight from Konstantin Wolf’s books – to print, fill in and tick off. When printed, each template starts on a new A4 page. Tip: you can select individual pages in the print dialogue.
This is an English translation. The books and the PDF are in German; the helplines in the help box are services in Germany, Austria and Switzerland.
From the book
AI and Children: the Parents’ Guide to ChatGPT & Co.
The seven templates from the guide. As of 8 October 2026. Explanations, step-by-step instructions and sources are in the book.
AI and children · Template 1
The 15-minute quick check for parents
- Which AI services and apps with AI does my child use? (ChatGPT, Gemini, Instagram/WhatsApp with Meta AI, Snapchat My AI, TikTok, Character.AI, voice assistant)
- Does my child’s age match the provider’s rules? (age limit table, chapter 3)
- Are parental controls set up for the most important services? (chapter 4)
- Are there quiet times – for example at night and at mealtimes?
- Do app downloads on the device need approval? (chapter 4.8)
- Have we talked about AI – and is there a family agreement? (chapter 5)
- Does my child know who to turn to if something feels strange? (chapter 6)
AI and children · Template 2
Settings checklist per service
ChatGPT
- Parental controls linked, invitation accepted
- Sensitive content reduced
- Model training off
- Memory off
- Voice mode / image generation / group chats: deliberately decided
- Quiet hours set
- Study mode / study times discussed
Google / Family Link
- Family Link set up
- App approvals
- Screen time and bedtime
- Filters for Play Store, Chrome and Search
Meta – Instagram, WhatsApp, Meta AI
- Parental supervision / Family Center set up
- Meta AI insights reviewed together
- Time limits and breaks
- WhatsApp parental controls (if available)
- Muse: not for children
Snapchat
- Family Center connected
- My AI disabled or deliberately allowed
TikTok
- Family Pairing connected
- Filter for AI-generated content checked
Alexa
- Amazon Kids activated
- Skills approved
- Time limits / bedtime
- Voice recording setting
Device
- App downloads only with approval
- Companion apps discussed
AI and children · Template 3a
AI family agreement for children (ages 6–12)
- We use AI together. I’m allowed to use these AI services:
- When: only with Mum/Dad/Grandma/Grandpa there at these times:
- What for: school fun and ideas
- I never type in: my name, my address, my school, photos, passwords, where I am right now.
- If the AI says something strange or scary: I stop and tell a grown-up straight away.
- AI is a tool, not a friend.
- My promise as the grown-up: I will listen, not scold, and not ban everything straight away.
- Child’s signature: Adult’s signature:
- We’ll look at this again on: (every 3 months)
AI and children · Template 3b
AI agreement for teens (ages 13–17)
- Services I use:
- Quiet times: at night from to · at mealtimes · other:
- What I use AI for: school (following my school’s rules, disclosing AI help) information (I check important facts) creative stuff and fun
- Never-type-this list: full name, address, school, photos of me and others, passwords, location.
- Parental controls: we chose these settings together:
- If something goes wrong (strange answers, pressure, fake images, a bot becoming too important): I’ll talk to or use the help box (template 5).
- My promise as a parent: to listen, not to ban things straight away, no secret monitoring, and to respect your privacy.
- Signatures:
- Review on: (every 3 months)
AI and children · Template 4
The never-type-this list for children
To hang up next to the computer.
STOP – this stays with me:
- My full name
- My address
- My school and class
- Photos of me, my family and my friends
- Passwords and PINs
- Where I am right now
If I’m not sure: ask first, then type.
AI and children · Template 5
Warning signs card and help box (to cut out)
Warning signs: withdrawal · use at night · bot as “only friend” · secrecy · drop in mood after chats · conversations about self-harm, food or weight.
What to do: bring it up calmly · listen · don’t take the device away immediately · get professional help.
Warning signs are not a diagnosis – they are a reason to talk.
HELP BOX (all details checked on 8 Oct 2026; services in Germany, Austria and Switzerland, mainly in German)
- Immediate danger to life: 112 (emergency services) / 110 (police, Germany)
- DE – Nummer gegen Kummer, children’s and youth helpline 116 111: free, anonymous, Mon–Sat 2–8 pm; does not appear on the phone bill
- DE – Elterntelefon (parents’ helpline) 0800 111 0 550: Mon/Wed/Fri 9 am–5 pm, Tue/Thu 9 am–7 pm, anonymous and free
- DE – TelefonSeelsorge 0800 111 0 111, 0800 111 0 222 or 116 123: around the clock, anonymous, free of charge
- AT – 147 Rat auf Draht: around the clock, anonymous, free; chat; parents’ site elternseite.at
- CH – 147 Pro Juventute (Beratung + Hilfe 147): around the clock, free, confidential; phone, WhatsApp chat via 147.ch, email beratung@147.ch
AI chatbots are not crisis support. In acute danger, call 112 immediately.
AI and children · Template 6
Homework cheat sheet
Five good learning prompts
- Explain: “Explain [topic] to me so that I can put it into my own words. Then ask me whether I’ve understood it.”
- Quiz me: “Make me a quiz with 5 questions on [material]. Only tell me after I answer whether I got it right.”
- Feedback: “Here is my own text. What’s good, and what can I improve? Don’t rewrite it for me.”
- Plan: “I have a test on [topic] in two weeks. Make me a study plan with 15 minutes a day.”
- Counter-arguments: “I think [opinion]. Give me three good counter-arguments so I can debate better.”
Three rules
- School rules apply. Ask your teacher what is allowed.
- Disclose AI help. Say so when an AI has helped you.
- Check the facts. 57% of young people consider AI information trustworthy (JIM study 2025) – but AI can be wrong. Look up important facts in a second source.
AI and children · Template 7
Parents’ evening plan (60 or 90 minutes)
| Time (60 / 90 min) | Part | Materials |
|---|
| 5 / 10 min | Opening poll by show of hands: “Who has tried ChatGPT? Whose child uses Snapchat, Instagram or WhatsApp?” | – |
| 10 / 15 min | Input from chapters 1 and 2: what children use AI for, the Common Sense test (fairly presented, with OpenAI’s counter-position), the five risk areas | Slide bullet points below |
| 15 / 25 min | Live demo of the settings from chapter 4 (with a test account – don’t show real children’s accounts) | Projector, template 2 |
| 20 / 30 min | Small groups: discuss and adapt the family agreement | Templates 3a/3b, 4 |
| 10 / 10 min | Wrap-up: questions, hand out the help box | Template 5 |
Slide bullet points: 1 · AI has arrived in everyday life (JIM study 2025: 91%) · 2 · Four uses: school, information, fun, “company” · 3 · The Common Sense test – results and counter-position · 4 · Five risk areas · 5 · Age limit table · 6 · Showing the settings · 7 · Guide rather than ban · 8 · Help box.
Invitation text (to adapt):
Dear parents, our children ask ChatGPT for help with homework, chat with Meta AI in WhatsApp or with “My AI” on Snapchat. What’s behind it, which settings help, and how do we talk about it as a family? We warmly invite you to a parents’ evening “AI and Children” on [date] at [time] in [place]. No prior knowledge needed – feel free to bring your smartphone. We look forward to seeing you! [Name, contact]
From the book
Using AI Safely – the Privacy Check for ChatGPT, Gemini & Co.
The quick fixes and the never-type-this traffic light from the guide. As of October 2026. The step-by-step instructions for ChatGPT, Gemini, Claude, Copilot & Co., the master checklist and the 30-day routine are in the full e-book.
Using AI safely · Checklist 1
5-minute quick fixes
- Training/model improvement switched off in every service I use
- List of shared links opened and anything unnecessary deleted
- Cookie banner: only necessary cookies allowed (also checked afterwards)
- Sensitive old chats deleted
- Resolved: sensitive topics only in temporary/incognito mode
- Careful with feedback (thumbs up/down) – it can share the entire chat
Using AI safely · Chapter 6
The never-type-this traffic light
A rule of thumb – not a legal classification.
Red – never type this in
Doesn’t belong in any AI chat on a personal account, not even in temporary mode:
- Passwords, PINs, TANs, recovery codes, two-factor codes
- Full credit card and bank details, online banking logins
- ID card, passport, tax ID or social security numbers, photos of ID documents
- Other people’s health data (child, partner, parents, patients) with names
- Customer, employee or client data with names, addresses or numbers
- Trade secrets, unpublished contracts, internal figures, source code with access keys
- Other people’s private messages, photos or voice messages that were entrusted to you
Yellow – only anonymised or in temporary mode
- Your own health questions – without your name or date of birth, worded as generally as possible
- Money questions – with rounded amounts, without account or contract numbers
- Conflicts, relationships, the workplace – replace names with roles (“my boss”, “person A”)
- Texts containing personal data (job application, reference, customer email) – redact them first
- Location-based questions – city instead of exact address
Green – safe
- General knowledge, explanations, study questions
- Recipes, travel tips (without travel dates and booking numbers), leisure ideas
- Texts that are going to be public anyway (blog post, product description)
- Ideas, outlines, brainstorming without confidential details
- Translations of public texts
The 3-second test before every send: “Would it be bad if this text turned up tomorrow as a chat title on an advertising platform or in a public link?” If yes: redact it or use temporary mode. If very much yes: don’t send it at all.
Using AI safely · Checklist 7
Traffic-light check before sending
- Does my text contain passwords, codes, bank details or ID numbers? → Stop
- Does it contain other people’s data? → redact or leave it out
- Is it about health, money or relationships? → temporary chat, anonymised
- Is a file attached? → open it first and check what’s really in it (including metadata)
- Do I want to share the chat? → then green content only
The redaction trick: before pasting, replace all names, addresses, numbers and company names with placeholders like [NAME], [COMPANY], [AMOUNT]. The AI can still help – and you put the real details back in yourself afterwards.