The “No Upload” Rule: Stop Kids from Sharing Personal Info with AI Chatbots

Why the “No Upload” rule matters (even when the chatbot feels like a friend)

AI chatbots are amazing for homework help, brainstorming stories, practicing languages, or even explaining a tough science concept. The problem is that kids (and adults) often treat them like a private diary or a trusted buddy. And that’s where families get into trouble.

Many kids don’t realize that what they type can be:

• Stored and reviewed to improve systems (depending on the service and settings)
• Shared unintentionally if they copy/paste the chat elsewhere
• Used to identify them when combined with other details (school name + city + photo = “easy to find”)

This is why “kids sharing personal information with ChatGPT” (or any chatbot) is one of the most common digital safety risks right now: it happens casually, not maliciously.

A strong family rule for AI privacy doesn’t have to be complicated. The simplest, stickiest rule we teach is:

The “No Upload” Rule: Don’t upload, paste, or type anything into an AI chatbot that you wouldn’t put on a public poster at school with your name on it.

This rule works because it avoids debates about which AI is “safe.” It builds a habit: treat chatbot prompts as public by default.

What not to tell AI chatbots (kid-friendly list + parent checklist)

When parents search “what not to tell ai chatbots kids,” they’re usually hoping for a clear list they can turn into a house rule. Here it is—written in plain language you can share with your child.

Kids should not share:

• Full name, home address, phone number, personal email
• School name, teacher names, team names, club names, schedules
• Passwords, passcodes, two-factor codes, login links
• Photos of themselves, their friends, or family (including selfies)
• Photos of homework with their name/school on it
• Medical info, diagnoses, therapy notes, medication details
• Private family info (custody details, finances, arguments, where parents work)
• Location details ("I’m home alone," “We’re at the airport,” “My soccer practice is at 6 at…")
• Anything they wouldn’t want read aloud in class

Two “sneaky” categories parents often miss:

• Identifiers inside files: A screenshot might show the school logo, a class code, a browser tab with a name, or a location in the background.
• Combination clues: A first name + city + school mascot + a coach’s name can identify a child even without an address.

Here’s a simple table you can put on the fridge (or in your Notes app) to make the rule concrete.

If your kid wants to share…	Risk (in plain terms)	Safer alternative prompt they can use	Parent tip
A selfie or photo of a friend	Face + background can identify them	“Describe how to improve lighting in a portrait photo without using my picture.”	Teach: “No faces in chatbots.”
A screenshot of homework	Name/class info may be visible	“Explain how to solve a problem like: 3/4 ÷ 1/2. Then give me 3 practice questions.”	Encourage re-typing the question without headers.
Their school name and grade	Easy to narrow down identity	“I’m a middle school student. Explain photosynthesis in simple steps.”	Use broad terms: “middle school,” “high school.”
A chat about bullying with names	Could expose others + escalate conflict	“How can someone respond to mean messages? Give options for getting help from a trusted adult.”	Keep it about actions, not people.
Passwords or login errors	Immediate account takeover risk	“What are common reasons a login fails? Give troubleshooting steps.”	Rule: “Passwords never leave your mouth or your fingers.”
Medical or therapy details	Very sensitive, long-term privacy	“What are general healthy ways to manage stress before a test?”	For personal health: talk to a real professional.

If you want one sentence that covers almost everything:

“No Upload means no faces, no names, no logins, no location, no documents.”

How to teach kids data privacy without scaring them (scripts that actually work)

Most kids don’t break privacy rules because they’re reckless—they break them because they’re focused on the task. A chatbot asks, “What school do you go to?” and it feels normal to answer.

Instead of fear-based lectures, use a quick, repeatable routine. This is the approach we recommend at Intellect Council for families learning how to teach kids data privacy.

1) Teach the “Pause-Check-Replace” habit (30 seconds)

Pause: “Before you hit send, pause.”

Check: “Does this include any of our No Upload items—names, faces, addresses, school, passwords, or documents?”

Replace: “Can you rephrase it more generally?”

Make it practical:

• Replace “I go to Lincoln Middle School in Denver” with “I’m in middle school.”
• Replace “Here’s my essay with my name at the top” with “Here’s the topic and my outline.”
• Replace “My password reset code is 839201” with “I’m not receiving a reset code—what should I check?”

2) Use a short family script (copy/paste ready)

Try one of these at the dinner table:

• For younger kids (5–9): “Chatbots are helpers, not secret keepers. We don’t give helpers our address, school, or pictures.”
• For tweens (10–12): “AI is like a public notebook. Don’t write anything you wouldn’t show your teacher.”
• For teens (13–17): “Even if an AI feels private, treat it like it could be saved or seen. Protect your future self.”

And when your child asks, “Why?” keep it simple:

• “Because privacy is power. Once something is shared, you can’t always take it back.”

3) Give kids a “safe prompt starter pack”

Kids share personal info because they don’t know another way to ask. Give them better defaults:

• “Explain this concept step-by-step using a made-up example.”
• “Quiz me with practice questions (no personal details).”
• “Help me brainstorm ideas for a story about a soccer team (fictional names).”
• “Summarize this topic at a 6th/9th-grade level.”
• “Help me write a polite message. Use placeholders like [TEACHER NAME].”

4) Make one rule for files: “If it’s a file, it’s private.”

A lot of “oops” moments come from uploads: PDFs, screenshots, photos, and documents.

Family rule you can say out loud:

• “If it’s a file, it’s private—don’t upload it to a chatbot without asking a parent.”

This single line prevents most high-risk sharing.

Setting up your home “AI Privacy Rulebook” (simple, enforceable, drama-free)

Rules fail when they’re vague. The best family rule for AI privacy is short, visible, and paired with a consequence that teaches (not punishes).

Here’s a practical rulebook you can adopt today.

The 4 family rules

• Rule 1: No Upload. No photos, screenshots, documents, or voice notes into chatbots without a parent check.
• Rule 2: No Identifiers. No full names, school names, usernames, addresses, phone numbers, or schedules.
• Rule 3: No Secrets. Don’t share passwords, codes, or private family information.
• Rule 4: Ask When Unsure. If it feels personal, pause and ask.

What “parent check” looks like (keep it quick)

Make it a 10-second routine:

• “Show me what you want to send.”
• Scan for: faces, names, school, location, logins, file headers.
• Help them rewrite it together.

A consequence that teaches

Pick one clear, calm response if the rule is broken:

• Delete the message (if possible), then rewrite safely.
• Take a 24-hour pause from chatbot use.
• Do a “privacy redo”: child explains what they shared and how to rephrase it.

The goal is skill-building: you’re training judgment.

Make privacy the default on devices (quick checklist)

• Turn on parental controls where available.
• Use kid-appropriate accounts when possible.
• Review chatbot settings for data sharing/training options.
• Keep chat history limited when the platform allows it.
• Teach kids to avoid linking accounts they don’t need (fewer connections = less exposure).

Even with good settings, the No Upload rule still matters—because the biggest risk is what kids type, paste, or attach.

Next Steps: How to get started tonight (10 minutes)

1. Have the 2-minute talk.

   • Say: “We’re starting a new rule for AI: No Upload and No Identifiers. Chatbots help, but they don’t need private info.”

2. Post the five “No Upload” items somewhere visible.

   • Names
   • Faces
   • Location
   • Logins
   • Documents

3. Practice with one real scenario.

   • Ask: “Show me how you’d ask a chatbot for homework help.”
   • Rewrite it together using placeholders like [SCHOOL] or [TEACHER].

4. Create a family “safe prompt” note.

   • Add 5–10 starters your child can copy/paste.

5. Do a weekly 5-minute check-in.

   • “Any weird chatbot questions this week?”
   • “Anything you weren’t sure about?”

If you do nothing else, do this: teach your child to replace personal details with placeholders. It keeps the learning benefits of AI while dramatically lowering the privacy risk.