Scams used to be easy to spot: bad grammar, strange email addresses, obviously fake offers. Artificial intelligence has changed that. Today a scammer can clone a loved one's voice from a short social media clip, create a realistic video of a public figure, or write a flawless email that looks exactly like your bank. The good news is that most AI scams still follow the same patterns. Once you know what to look for, you can stop them before they cost you money or personal information.
Why AI scams are different
AI tools let criminals do three things faster and more convincingly than before: impersonate people you trust, personalise messages using details found online, and scale the same trick to thousands of people at once. That means you can no longer rely on spelling mistakes or an unfamiliar voice to warn you. Instead, focus on what the message is asking you to do.
9 red flags of an AI scam
- Urgency and panic. "Act now", "your account will be closed", "I'm in trouble, please hurry." Pressure is designed to stop you thinking.
- A familiar voice asking for money. A call from a "grandchild" or "boss" who needs an urgent transfer is a classic voice-clone scam.
- Requests for secrecy. "Don't tell Mum" or "keep this confidential" isolates you from people who could spot the trick.
- Unusual payment methods. Gift cards, crypto, wire transfers or payment apps to a new account are almost always a warning sign.
- Video that looks slightly off. Blurry edges around the face, odd blinking, lips that don't quite match the words, or lighting that changes strangely.
- Too-good-to-be-true investments. Guaranteed returns, "AI trading bots" and celebrity endorsements you can't verify elsewhere.
- Links and login pages you didn't ask for. Even a perfect-looking email can lead to a fake login page. Go to the website yourself instead.
- Job offers that skip the interview. Instant hiring, chat-only interviews, or requests to pay for equipment up front.
- Requests for codes. No genuine bank, company or relative needs your one-time passcode. Sharing it hands over your account.
A simple family safety plan
- Create a family code word. Agree on a private word that anyone can ask for during an "emergency" call. A cloned voice won't know it.
- Pause and verify. Hang up and call the person or company back on a number you already have.
- Lock down what's public. Limit voice and video clips, birthdays and family details on public social media profiles.
- Turn on two-factor authentication for email, banking and social accounts.
- Talk about it. Share examples with older relatives and teenagers — they are among the most targeted groups.
What to do if you think you've been scammed
Act quickly. Contact your bank to stop or reverse payments, change passwords on any affected accounts, and report the scam to your country's fraud-reporting service (for example, the FTC at reportfraud.ftc.gov in the US). The faster you move, the more you can usually recover.
Go deeper
If you'd like step-by-step checklists and scripts, these guides cover each situation in detail:
- AI Scam Awareness for Seniors — a clear guide for older adults and the people who care for them.
- The AI Phishing Response Plan — exactly what to do in the first hours after a scam.
- The Deepfake Defense Protocol — protect your face, voice and identity from AI mimicry.
- AI-Proofing Your Online Banking — lock down your accounts and payments.
Browse the full Digital Safety & Security collection for more.