Why Safety Matters More Than Ever
AI makes scammers more dangerous. In the past, you could spot a scam email because of bad grammar or obvious mistakes. Now, scammers use AI to write perfectly polished, convincing messages — often using your name and personal details they've found online.
The good news: the warning signs haven't changed. What scammers want from you is still the same. If you know what to look for, you can protect yourself.
The Three Big Risks
🚨 Risk 1: AI-Powered Scams
Scammers are now using AI to make their schemes more convincing than ever. They can:
- Write flawless emails pretending to be your bank, Medicare, or the IRS
- Clone voices — using a short audio clip from social media to fake a call from your grandchild saying they're in trouble
- Create fake websites that look exactly like real ones
- Generate fake "government documents" or invoices
The "Grandparent Scam" — Now Powered by AI: You get a call. It sounds exactly like your grandchild's voice, saying "Grandma, I'm in jail — please send bail money right away and don't tell Mom." This call may be AI-generated from just a few seconds of real audio. Always hang up and call your grandchild's real number to verify.
The ONE Rule to Remember
Stop when an unexpected caller pressures you to send money urgently. Demands for gift cards, cryptocurrency, or an immediate transfer are warning signs. End the call and independently verify the request through a trusted number before paying.
Make a plan with your family
A family safe word: Pause. Ask. Call back.
AI can imitate a familiar voice. If a distressed caller sounds like your child, grandchild, or another loved one, check who is calling before acting on a request for money or private information.
- Pause. Take a breath. Urgency, secrecy, and a familiar voice are not proof of identity.
- Ask. Ask the caller to tell you your family's private code word or phrase. Do not say it first or offer hints.
- Call back. End the call and dial a saved, trusted number. Do not use a new number supplied by the caller. Even a correct phrase is only one check.
Choose your phrase before an emergency
Agree privately on something memorable that is not a birthday, pet name, or public fact. Share it only with trusted family members and replace it if it becomes known to others. Never enter the actual phrase into an AI tool or this website, and do not write it on your fridge card.
Cannot verify the caller? Contact another trusted person. Not knowing the phrase does not establish whether someone is safe or in danger. If you believe there is immediate danger, call your local emergency number; in the U.S., call 911.
Practice together
Use a pretend phrase for a short role-play. Practice pausing, asking, and finding the trusted number you would call. Do not make real emergency calls during practice.
Download the family safety fridge card (PDF)
Sources: FTC family emergency scam guidance and FBI family code word guidance. Checked September 2026.
📰 Risk 2: AI Misinformation
AI can generate convincing-sounding articles, social media posts, photos, and videos that contain false information. This is sometimes called "deepfakes" (for videos) or "AI hallucinations" (for text that AI made up).
- AI-generated photos can show events that never happened
- Fake "news articles" can appear on real-looking websites
- Health misinformation spreads fast on social media — AI makes it worse
- Even AI chatbots can make up facts and present them confidently
🔐 Risk 3: Your Privacy
When you type something into an AI tool, that information may be stored by the company. This isn't always a problem, but it means you should be thoughtful about what you share.
- AI companies may use your conversations to improve their models
- Information you share is stored on their servers — treat it like a postcard, not a sealed letter
- Third-party AI tools (not the major ones) may have unknown privacy practices
Your Personal Safety Rules
Hang up and call back yourself
If you get an unexpected call from your bank, Medicare, the IRS, or a family member asking for money — hang up. Find the real phone number yourself and call back. Scammers rely on the pressure of the moment.
Never pay with gift cards
Gift cards (like Google Play, Amazon, or iTunes cards) are untraceable cash. No real government agency, utility company, or family member will ask you to pay this way.
Pause before you share
Before sharing a news story, a photo, or a health claim on social media — pause for 30 seconds. Ask: "Do I know for sure this is true?" If not, don't share it.
Keep personal information personal
Never type your Social Security number, full Medicare ID, bank account numbers, or passwords into any AI tool — or any website you didn't navigate to yourself.
Talk to someone you trust first
If you receive an alarming message, an unusual request, or an offer that seems too good to be true — call a family member, close friend, or trusted neighbor before doing anything. Scammers want you to act alone and fast.
Verify health information
AI can provide general health information, but it can also be wrong. Always verify with your doctor, pharmacist, or trusted sites like mayoclinic.org or nih.gov before changing your medications or treatments.
🛑 If You Think You've Been Scammed
Don't be embarrassed — it happens to very smart people. Act quickly:
1. Call your bank immediately to freeze your accounts.
2. Report to the FTC at reportfraud.ftc.gov
3. Tell a trusted family member or friend right away.
4. If you gave gift cards, call the card company — sometimes they can freeze funds.
Safety Takeaway for This Lesson
- No government agency ever asks for gift card payments — ever
- AI can clone voices — always call back on a number you look up yourself
- Don't share personal information in any AI tool
- Pause before sharing anything online — misinformation spreads fast
- If something feels wrong, trust that feeling and talk to someone before acting
Make it yours
My Family Safety Plan
Practice responding to an urgent family call using a pretend code word. Later, agree on your real family phrase privately and complete your trusted callback contacts.