AI Scams: How to Spot Deepfakes and Voice Cloning Fraud
It used to be easy to spot a scammer: broken language and typos. That era is over. AI can now fake a person’s voice, face, and writing style — and scams have become far more convincing.
The good news: protecting yourself isn’t hard. Every AI scam relies on the same two levers — urgency and emotion. Once you recognize them, you’re halfway safe.
How scams have changed
Voice. A few seconds of audio is enough for AI to recreate a person’s speaking voice. A scammer can call as your “grandchild” and ask for money in an emergency. The voice sounds real; the number is unknown.
Video. Deepfake videos can make any of us appear to do and say anything. The most common version spreads on social media: a “famous investor” or “news anchor” promising guaranteed returns. Bosses appear too: a company executive “asks” you to buy gift cards.
Text. Scam messages are now grammatically flawless. The bank’s, tax office’s, or postal service’s name and logo check out. Perfect writing is no longer a sign of a genuine message.
Warning signs
- Urgency and pressure. “Act now or lose everything.” Legitimate organizations give you time to think.
- Odd payment method. A bank transfer during the call, cryptocurrency, or gift cards.
- Secrecy. “Don’t tell anyone.” Real matters never require silence.
- Codes and passwords. Your bank will never ask for your PINs over the phone.
- Strange voice. Unnatural pauses, a metallic tone, or a commanding voice that pushes you to hurry.
How to protect yourself and your family
- Never make money decisions during a call. Never. Hang up and think.
- Verify on another channel. If a “relative” calls, call back on the number you already have for them.
- Agree on a family code word. Pick a shared password or question only your family knows the answer to. It stops a voice scam in seconds.
- Talk to your parents. Older people are the number one target — one conversation can save thousands.
- Report immediately. If you were scammed, call your bank right away and file a police report. Speed matters.
Read more: What is AI? and AI and Privacy
Frequently asked questions
Does three seconds of audio really suffice to fake a voice?↓
Unfortunately, yes. Even a short clip from a video or phone call is enough for a convincing voice clone today. That's why a voice is no longer proof of identity — always verify with a separate call to a number you already have.
What should I do if I already transferred money to a scammer?↓
Call your bank immediately and ask them to stop the payment. Then file a police report — in most countries this can be done online. The faster you act, the better your chances of getting the money back.
What should I do if I see a deepfake video on social media?↓
Report the video to the platform (for example Facebook or TikTok) using their report function. If it impersonates a public figure, consider warning the person or their employer. Don't share it onward — sharing helps the scammer, even with good intentions.
Related articles
AI and personal finance — how it actually helps with everyday budgeting
Where does a chatbot really fit in managing your money, and where does it not? Practical examples of budgeting, weighing up big purchases and saving — without ever handing over your bank credentials.
AI and Maps — How Does Google Maps Know the Road Is Blocked
Traffic predictions, arrival times, and route suggestions in map apps are AI. We explain where the data comes from and why the forecast sometimes fails.
AI in Customer Service — Why a Bot Answers the Phone and How to Reach a Human
Banks, shops, and public agencies increasingly use AI bots in customer service. How to recognize a bot, when it's useful, and how to ask for a real person.