How to Spot AI Scams and Fake Messages: A Self-Defense Guide for Everyone

AI is genuinely useful, but the bad guys are using it too. Scams and fake messages have gotten a lot harder to spot in the last few years, and AI is the reason why. The good news: you don't need to understand the technology. A few simple habits will block most of it. Everything in this guide works for anyone, and it's worth passing along to the older folks in your family.

What AI scams look like now

Compared to a few years ago, today's scams are far more convincing. Here are the common ones.

Fake voices. Scammers use AI to clone the voice of someone in your family and call you, sounding genuinely like them, claiming they're in trouble and need money fast. All it takes is a short clip of that person talking somewhere online, and they can copy the voice in seconds.

Fake videos and fake faces. They splice a celebrity's face, or the face of someone you actually know, into a video that person never recorded, then use it to push you toward an investment or get you to click a link. These days even a video call can be fake.

Fake support lines and fake official notices. They impersonate your bank, your phone company, a shopping site, or a government agency, and build a website or page that looks almost identical to the real thing, all to trick you into entering your username and password.

Fake influencers and fake investments. They use AI to conjure up a polished-looking "teacher" or "analyst" who pulls you into a chat group, promises a sure-thing return, and then walks off with your money.

Why this is so hard to tell apart now

We used to judge what was real by how it looked and how it sounded. AI broke exactly that test: it can make fake things look and sound just like the real thing.

So the old rules ("are there typos?", "is the picture blurry?") don't really work anymore. You need a different approach: don't ask whether it looks real, ask whether it makes sense and whether it's rushing you. The giveaway in a scam is almost never in the image or the audio. It's in what they're asking you to do.

A few checks that need no tech skills at all

Remember these, and you'll be safer than anyone who's memorized a list of detection tricks.

If it's urgent, stop first. Scams almost always manufacture a sense of urgency: limited time, right now, it'll be too late if you don't act. Real institutions and real family members rarely pressure you to transfer money or hand over information "this instant." Urgency itself is the signal to slow down.

If it involves money or a password, be very suspicious. No matter who they claim to be or how convincing they sound, the moment the request is to transfer money, share a verification code, enter your username and password, or click a link to log in somewhere, treat it as a scam until proven otherwise.

Verify through a different channel. If you get a call saying a family member is in trouble, hang up and call that person back yourself on the number you already have. If you get a "notice from your bank," don't tap the link in the message: call the bank's own support line or open their official app. Anything genuine will hold up when you confirm it a second way.

Ask a question only the real person would know. If you suspect the voice on the phone is fake, ask about something only the real person could answer. AI can copy a voice, but it can't answer the private things that only the two of you share.

Treat links and unexpected calls as off-limits by default. Don't tap links from strangers, and don't do what an unknown caller tells you to. Better to be a little slow and a little inconvenienced than to rush along with someone else's instructions.

Your standard move when a suspicious message arrives

You don't have to think it through every time. Just keep one fixed routine.

1. Stop. However urgent it feels, put the phone down for three minutes.
2. Don't click, don't hand anything over. No links, no passwords or codes, no immediate transfers.
3. Verify it yourself. Use a number you already have or the official app, and reach out to that institution or that person on your own.
4. When in doubt, ask. Call a fraud-reporting hotline, or ask someone in the family who's comfortable with the internet. There's no shame in asking. Getting scammed is the real headache.

How to remind older relatives and family

Older folks aren't slower on the uptake. These new scams catch younger people too. When you bring it up, skip the technical lecture and just hand them the simplest line: "Urgent, wants money, wants a password? Hang up and call back yourself to confirm."

You can also agree on a "code word" with your family ahead of time, or a question only your household knows the answer to. That way, if a call comes in saying "It's me, your son, I'm in trouble," one question settles whether it's real. Agreeing on this in advance beats trying to fix things after the fact, by a long way.

The one-line takeaway

AI makes fake things look real, so stop judging by whether it looks real. Judge by whether it's rushing you and whether it wants money or a password. The moment someone is pressuring you, asking for money, or asking for a password, stop, switch to a different channel, and confirm it yourself. Build that habit, teach it to your family, and most AI scams won't get anywhere near you.