Helping Kids Stay Safe Online & With AI

You do not need to understand every app your child uses. You do need to establish a few consistent expectations:

• Pause before clicking, replying or paying.
• Keep sensitive information private.
• Use strong passwords and MFA.
• Verify surprising information.
• Treat AI output as something to evaluate, not automatically trust.
• Ask for help early.

A useful message to establish before an incident occurs is: If something goes wrong online, tell me. We will deal with the problem first.

Children are far more likely to tell an adult about an online problem if asking for help does not automatically mean losing access to every device.

A single behavior change usually does not prove that something dangerous is happening. Look for patterns.

• Sudden secrecy around online activity
• Panic or intense anxiety after receiving messages
• A new online relationship the child will not discuss
• Someone repeatedly moving the child from a public platform into private messaging
• Unexpected requests for money, gift cards or crypto
• Requests for verification codes
• Secret accounts or sudden deletion of conversations
• Threats involving photos or screenshots
• Major changes in mood after being online
• Increasing emotional dependence on an AI chatbot or unknown online person

The goal is not surveillance for its own sake. The goal is noticing when normal digital activity begins to include pressure, secrecy, fear or loss of control.

One of the most useful cybersecurity lessons has very little to do with technology.

Social engineering works by influencing a person's behavior. An attacker may create:

• Urgency: "You have five minutes."
• Fear: "Your account will be closed."
• Secrecy: "Don't tell your parents."
• Flattery: "You're so mature. I can trust you."
• Reward: "You've won."
• Romantic attention: "No one understands me like you do."
• Isolation: "Other people are trying to keep us apart."

Children should know: If somebody is trying to control how quickly you act, who you tell or how you feel, pause.

Generative AI has changed one of the basic assumptions many adults grew up with. Seeing a photo is no longer strong proof that something happened. Hearing a voice is no longer strong proof of identity.

A screenshot can be fabricated. A video can be altered. A person can create a convincing profile using content that belongs to somebody else.

Teach this habit: Important + surprising + urgent = verify another way.

If a relative appears to ask for emergency money, call that relative using the number you already have. If a school login request appears unexpectedly, open the school's normal website instead of following the message link.

A blanket "don't use AI" rule is becoming increasingly difficult to enforce and may prevent useful conversations. Instead, teach children what responsible AI use looks like.

AI can help with brainstorming, explaining difficult ideas, practicing questions, organizing information, generating creative ideas and learning new skills.

But children should understand that AI can also:

• Generate incorrect information
• Confidently state invented information
• Reproduce bias
• Create realistic fake media
• Collect information the user enters
• Imitate human conversation without actually being human

California's current guidance for public schools addresses data privacy, academic integrity and human-centered AI use, while UNESCO emphasizes critical evaluation, ethics and human agency as part of AI competency. UNICEF's updated AI-and-children guidance also places child safety, privacy, transparency and well-being among the central requirements for child-centered AI.

AI can sound remarkably patient, confident, supportive and personal. A young person may begin treating a chatbot as if it has human understanding, judgment or loyalty.

Adults should pay attention if a child begins:

• Preferring an AI relationship to human relationships
• Hiding conversations with an AI system
• Relying on AI for major emotional or personal decisions
• Assuming the system "knows" or "cares about" them
• Sharing increasingly sensitive information
• Treating AI-generated advice as authoritative

The right response is usually conversation before confrontation. Ask: "What do you like about talking to it?" "Does it ever say things that seem wrong?" "How would we know whether its advice is reliable?"

Children need age-appropriate privacy. They also need to understand that digital privacy has limits.

Teach them never to casually share:

• Full legal names
• Home addresses
• School schedules
• Live location
• Passwords
• Verification codes
• Private documents
• Financial information
• Intimate images
• Sensitive family information

The same principle applies to AI tools. Before entering information, ask: Does this system actually need to know this? If not, leave it out.

Once children begin using debit cards, Apple Pay, Venmo or other payment systems, financial safety should become part of digital-safety education.

• Never send a one-time code.
• Never let someone temporarily use your account.
• Never move money for an online acquaintance.
• Be skeptical of gift-card or crypto payment requests.
• Verify payment requests separately.
• Check the real account rather than trusting a screenshot.

A first payment card should come with a short cybersecurity conversation just as naturally as it comes with a conversation about budgeting.

The first response matters. Try starting with:

• "Thank you for telling me."
• "Show me what happened."
• "We will figure out what to do next."

Avoid beginning with: "Why did you do that?" "I told you so." "I'm taking your phone." Your first goal is to understand whether there is an active threat and preserve useful information.

• Save screenshots or other evidence.
• Stop communication with the suspicious account.
• Block and report.
• Secure compromised accounts.
• Check financial activity.
• Inform the school if other students are involved.
• Escalate serious threats, exploitation or criminal behavior to the appropriate platform or authorities.

FBI and NCMEC both provide dedicated information and reporting resources for child online exploitation and sextortion.

Students may tell a teacher something they have not told a parent. Schools can help by making the next step predictable.

Students should know:

• Which adult they can approach
• How to report something privately
• What happens after a report
• When confidentiality cannot be maintained
• Where to find school-approved AI guidance
• How academic AI use differs from unsafe or prohibited AI use

Digital safety is easier when the rules are clear before an incident occurs.