How to Spot AI Fake Voices, Photos and Messages
Plain checks for AI-made voices, photos and messages: what to look for, why it is getting harder, and a family code word plan that works even when you cannot tell.
You often cannot tell by looking or listening alone, so do not rely on your eyes and ears. Slow down, check through a second channel you already trust (call the person back on a saved number), and agree on a family code word for emergencies.
AI can now copy a voice from a few seconds of audio, create realistic photos, and write messages with perfect spelling. That does not mean you are helpless. It means the old rule, “look for typos,” is no longer enough.
This guide is general safety information, not legal advice. If you have lost money, contact your bank and report it at the end of this guide.
Why this is getting harder
Older scams were easy to spot: poor grammar, strange links, odd requests. AI removes many of those signs. A scammer can now write a smooth message, clone a voice from a video posted online, or make an image that looks like a real photo.
What AI cannot easily fake is a process. A scammer can sound like your grandson, but cannot easily pass a code word you set up, call back from the real number, or show up in person. So the best defenses are habits, not detective skills.
Fake voices: the “emergency call”
The classic version is a call from someone who sounds like a family member. They say they are in trouble, in jail, in a crash, or stranded. They ask for money fast, often by gift cards, wire transfer or cryptocurrency, and ask you not to tell anyone.
Signs to watch for:
- Urgency, panic and secrecy
- A request for unusual payment such as gift cards or crypto
- A call from an unknown number, or a number that looks familiar but is spoofed
- Odd pauses, flat emotion, or a voice that sounds slightly robotic
- The caller avoids answering personal questions
What to do:
- Hang up. You can be polite and still end the call.
- Call the person back on the number saved in your phone.
- Ask for the family code word. If they cannot give it, stop.
- Call another relative to check the story.
- Do not send money until you have confirmed it through a second channel.
Fake photos and videos
AI-made images are common on social media. They can be funny and harmless, or used to sell something or spread a false story.
Look for clues, but treat them as hints, not proof:
| Clue | What to look for |
|---|---|
| Hands and teeth | Extra fingers, odd shapes, too many teeth |
| Text in the image | Garbled letters on signs or labels |
| Backgrounds | Warped lines, objects that blend together |
| Skin and hair | Too smooth, or hair that merges with the background |
| Lighting | Shadows that do not match |
| Context | No source, no date, no original poster |
The clues are fading as tools improve. A better habit is to check the source.
- Who posted this first?
- Is a trusted news source reporting it?
- Does a reverse image search (in your browser’s image search) find an earlier version?
- Does the post ask you to react emotionally, share quickly or donate?
If you cannot confirm it, do not share it.
Fake messages: texts, emails and chats
Scam messages are now polished. They may use your name, a real company logo and correct grammar. The real tell is usually what they ask you to do.
Common patterns:
- “Your account is locked. Click here to verify.”
- “You have an unpaid toll or package fee.”
- “This is your boss. I need gift cards today.”
- “I found your resume. Message me on this app for a job.”
- A friend’s account asks for urgent money after being hacked.
A message is suspicious when it:
- Creates urgency
- Asks for money, a password or a code
- Asks you to click a link or open a file
- Comes from an unexpected source
- Asks you to keep it secret
Do not click. Go to the company’s website by typing the address yourself, or open the official app, and check there.
A quick test: the S.T.O.P. check
Use these four steps before you act on anything unexpected.
- Slow down. Urgency is the scammer’s tool.
- Think about what is being asked. Money, codes and passwords are red flags.
- Opt for another channel. Call back on a saved number or visit the official site.
- Pause before paying. Never pay with gift cards, wire transfers or crypto for an unexpected request.
Set up a family code word
A code word is a simple, strong defense. Do it tonight.
- Choose a word or phrase that is not guessable and is not on social media. Avoid pet names and birthdays.
- Agree that anyone who calls for money or help must say it.
- Tell every family member, including teens and grandparents.
- Change it if it is ever shared.
Say it plainly: “If anyone ever calls asking for help, ask them for the code word. If they cannot give it, hang up and call me.”
Protect your own voice and photos
You cannot remove all risk, but you can reduce it.
- Set social media accounts to private.
- Do not post long videos of children speaking, with names and school details.
- Use a voicemail greeting that does not say your full name.
- Turn on two-step verification for your email and bank.
- Use a password manager.
What to do if you think you were scammed
- Stop all contact with the scammer.
- Call your bank or card company right away. The sooner you act, the better the chance of stopping a payment.
- Change passwords on any account you shared, starting with email.
- Report it. In the US, you can report scams to the Federal Trade Commission at ReportFraud.ftc.gov and to the FBI’s Internet Crime Complaint Center at ic3.gov.
- Tell someone you trust. Scams are common, and many smart people are fooled.
When to get help
- You sent money or shared a code or password.
- A scammer is threatening you, or someone you know.
- You feel distressed, ashamed or hopeless after being targeted. Talk to someone you trust. In the US, you can call or text 988 at any time if you are in emotional crisis.
- A fake image or video of you or your child is being shared. Contact the platform and local police, and keep screenshots.
A checklist to keep
- Set a family code word
- Save key family phone numbers in your phone
- Turn on two-step verification
- Never pay with gift cards for an unexpected request
- Call back on a saved number
- Check the source before sharing an image
- Tell older relatives about voice cloning
Talk to older relatives without scaring them
Parents and grandparents are often the target of voice-clone calls. A calm, short talk works better than a long warning.
- Start with a story: “A friend got a call that sounded like her grandson, and it was fake.”
- Explain the one habit: hang up and call back on a saved number.
- Set the family code word together, and write it on a card by the phone.
- Practice once. Call them, play the part of a scammer, and let them hang up.
- Remind them that nobody will be angry if they check, even if it turns out to be real.
Many people feel embarrassed about being fooled. Say plainly that these scams work on smart, careful people, which makes it easier for them to tell you if something happens.
A gentle next step
If you would like the plan on paper, Spotting AI Fakes: Voices, Photos and Messages walks through the checks above with examples. For a free place to start, try the free Spot a fake message page.
Tonight, text your family and pick a code word. It takes two minutes and covers a lot.
Products that go with this
Keep going
- Guide9 min readHow to Scam-Proof a Parent's Phone in One Afternoon
Settings to change, a family code word, and a simple pause-and-call-back rule, set up together in about three hours without lecturing.
- Guide9 min readHow to Make a Family Emergency Plan in One Evening
A two-hour plan for your household: who to call, where to meet, what to grab, and how to practice it so everyone knows what to do.
- Guide9 min readHow to Talk to Kids About AI Safely
Plain words and simple rules for talking with kids about chatbots, fake images and homework help, by age, with a family agreement you can set up in one evening.

