I recently found myself thinking about all the things I want my kids to know before they head out into the world on their own. As parents, we spend years teaching our kids the basics: lock the door, don’t give strangers your house key, be thoughtful about your finances, and be careful about who you trust with personal information. But there’s a new version of that conversation that we need to have—and it involves AI.
Whether you use ChatGPT, Gemini, Grok or another AI tool, these systems can be incredibly useful. AI can help you brainstorm dinner ideas, explain a complicated topic, write an email, plan a trip, or even help you figure out why your houseplant looks so unhappy. But because chatting with AI can feel so easy and conversational, it’s also easy to forget that you’re sharing information with a technology service—not whispering into the ear of a trusted friend.
And that’s where the problem lies. AI operates on vast amounts of processed information and then uses it to create responses to your questions. The information that you provide therefore may be retained in logs or conversation history, could be seen by employees working on the AI models or analytics, or could potentially be hacked or compromised. In other words, some of the same common-sense rules we teach our kids about locking the door or not leaving a bank card lying around apply to AI, too.
All this got me thinking about some basic rules I want my own kids to remember. And since they’re good rules for all of us, I thought I’d share them with you, too.
1. Never Share Your Passwords
This one should be a no-brainer, but it’s worth saying anyway.
Never give an AI your passwords, PINs, security codes, recovery codes, or other login credentials.
You might be tempted to say, “Here’s my password—can you tell me if it’s strong?” Don’t.
You can ask AI how to create a strong password or to explain what makes a password secure, without giving it an actual password you use.
2. Keep Personal Information Personal
Think twice before entering information that could identify you or someone else.
That includes things like:
- Detailed location information
- Private documents
- Exact birth dates
- Home addresses and phone numbers
- Personal email addresses
- Social Security numbers and other government ID numbers
- Passport or driver’s license information
- Bank and credit card numbers
- Account numbers
Some of these may seem harmless on their own, but pieces of information can be combined to create a surprisingly detailed picture of a person. National Institute of Standards and Technology (NIST) notes that AI systems can create privacy risks by connecting information from different sources or inferring sensitive information.
A good rule for kids—and grown-ups—is simple: If you wouldn’t hand it to a stranger, don’t paste it into a chatbot.
3. Don’t Share Someone Else’s Private Information
Your privacy isn’t the only privacy you need to protect.
Maybe your friend sends you a private message and you want help figuring out how to respond. Maybe a family member gives you a document to review. Maybe your coworker sends you an email containing their phone number.
Before putting any of it into an AI tool, stop and ask yourself: Do I have permission to share this?
The fact that information is available to you doesn’t mean it’s yours to upload.
This is especially important with children’s information, photographs, personal conversations, medical information, and anything someone shared with you in confidence.
4. Be Very Careful With Photos
AI can do amazing things with images. It can edit them, describe them, organize them, and transform them in all sorts of creative ways.
But before uploading a photo, take a second look at what’s actually in it.
Is there a child in the picture? A driver’s license sitting on the counter? A boarding pass? A school uniform? Your house number? A computer screen showing private information?
Even an innocent-looking family photo can contain more information than you realize.
And if the photo belongs to someone else, ask permission before uploading it.
5. Keep Your Financial Information Out of the Chat
AI can certainly help explain financial concepts. It can help you understand what a mortgage term means or make a list of questions to ask your financial advisor.
But there is a big difference between asking, “What is an adjustable-rate mortgage?” and pasting your bank statement into a chatbot.
Don’t upload:
- Bank statements
- Credit card statements
- Tax returns
- Investment account information
- Credit reports
- Pay stubs
- Loan documents containing account numbers
- Detailed financial records
If you want AI to help analyze something, remove identifying information first—or better yet, use fictional or generalized numbers.
6. Think Twice Before Sharing Medical or Other Sensitive Information
Health information can be particularly personal.
You may want AI to help you understand a medical term or prepare questions for your doctor, and there are perfectly reasonable ways to use AI as an educational tool. But avoid casually pasting medical records, insurance information, prescription details, or another person’s private health information into a chatbot.
The same goes for other sensitive areas of your life—legal problems, family disputes, relationship issues, or anything else you would consider deeply private.
AI can be a useful tool without becoming your personal diary.
7. Keep Business Secrets Out of AI
This is a big one for anyone who works from home, runs a small business, or uses AI at work.
Don’t paste confidential company information into a consumer AI tool just because you’re trying to save a few minutes.
That might include:
- Customer lists
- Private customer information
- Contracts
- Pricing strategies
- Unreleased products
- Business plans
- Financial statements
- Internal emails
- Trade secrets
- Proprietary code
- Confidential employee information
The Federal Trade Commission has specifically warned about the risks of businesses and users putting sensitive or confidential information—including internal documents and customer data—into AI systems.
If your employer has approved a particular AI system for work, follow the company’s rules. If you’re not sure whether something is okay to upload, ask before you paste.
8. Don’t Assume “It’s Just a Chat” – Understand AI Privacy Before You Share
This may be the most important lesson of all.
AI is designed to feel conversational. You type something, and it responds immediately. That can make it feel a little like texting a really knowledgeable friend.
But it’s still a technology service, and different AI products have different privacy policies, settings, retention practices, and ways of handling user information.
That’s why you should understand what you’re using before sharing anything sensitive.
NIST identifies privacy as an important risk area for generative AI, including the possibility of sensitive information being exposed, inferred, or leaked.
And this is a good lesson for kids, too: something that feels private isn’t necessarily private.
9. Don’t Upload Something Just Because AI Asks
AI can ask follow-up questions that sound completely reasonable.
It might say, “Upload the document and I’ll take a look.”
Or, “Send me a screenshot so I can help.”
Or, “Give me your account information and I can troubleshoot this.”
That doesn’t mean you should.
Before sharing anything, ask:
Does the AI actually need this information to answer my question?
And am I comfortable providing this information?
If the answer is no, leave it out.
In fact, this is a great habit for the entire internet—not just AI. The less sensitive information you share, the less sensitive information there is to lose.
10. When in Doubt, Remove the Details
You don’t have to stop using AI to protect your privacy.
You just need to get better at sanitizing what you share.
Instead of:
“Here’s my employee’s performance review. Can you rewrite it?”
Try:
“I’m writing a performance review for an employee. Can you help me make this feedback more constructive and professional?”
Instead of uploading your actual financial statement, describe the situation using made-up numbers.
Instead of sharing a private email with someone’s name, phone number, and address, remove those details first.
You can often get almost the same help without handing over the information you were trying to protect in the first place.
My Rule for the Kids—and for Me
When I was thinking about how to explain all of this to my kids, I realized I didn’t want to give them a complicated list of technical rules.
I wanted to give them one simple question:
“Would you be comfortable if this information somehow became public?”
If the answer is no, don’t share it with AI.
What Can You Safely Share with AI
That doesn’t mean AI is bad. Quite the opposite. I use technology because it can be incredibly helpful, and I want my kids to learn how to use these tools confidently and responsibly.
You don’t have to treat every conversation with AI as dangerous. The trick is knowing the difference between information that’s useful and information that’s sensitive.
It’s generally fine to ask AI to explain a recipe, brainstorm vacation ideas, help you rewrite a generic email, explain a concept, or create a fictional example. The more personal, confidential or identifying the information becomes, the more careful you should be.
When possible, remove names, account numbers, addresses and other identifying details before sharing something with an AI tool.
–
So use AI. Ask it questions. Let it help you brainstorm, learn, create, plan, and solve problems.
Just remember that not everything belongs in the chat box.
And honestly, that’s advice I think all of us could use a little reminder about from time to time.