As many of my younger friends would say, I’m an “Unc” — I remember when scams looked very different.
My Yahoo inbox used to be flooded with emails from so-called “international dignitaries” promising huge payouts if I just sent a small fee first. I’d pick up the landline and hear someone asking for my credit card details. It was obvious enough to ignore — but I’ve seen older family members fall for those exact tricks.
Now, I worry about those same people even more.
AI-powered scam calls are on another level. With voice cloning and deepfake audio, scammers can mimic someone you know — a family member, a friend — and make it sound frighteningly real. It’s no surprise that people who aren’t familiar with how advanced this tech has become are getting caught off guard.
Ironically, I turned to ChatGPT to understand how to spot these scams. The advice it gave me is simple, practical and something anyone can use to avoid falling victim to these increasingly convincing calls.
The red and subtle flags you need to keep an ear out for
After asking ChatGPT, “How can someone tell if a phone call is an AI scam?” it initially gave me a list of the most obvious signs that I made sure to keep in the back of my head for future reference:
- Urgency + pressure tactics: Scammers push you to act immediately (“send money now,” “don’t tell anyone”). Real organizations rarely operate like this.
- Requests for money or sensitive info: Especially via unusual methods like gift cards, crypto, or wire transfers.
- “Familiar” voice but strange context: AI can mimic someone you know, but the situation may feel off (“I’m in jail, don’t call back, just send money”).
- Refusal to let you verify: They may insist you stay on the line or avoid calling others.
Even though I consider myself tech-savvy and always trust my instincts when I smell a scam coming my way, the tips ChatGPT listed helped me become more aware of some AI scam call behaviors I never knew of.
Anyone who calls you and badgers you for delicate info connected to your financial records is a clear sign of a scammer. But it’s worth knowing that if a person on the other end of the phone refuses to let you call the person they’re saying is in danger, you’re probably in the middle of an AI scam call.
As a follow-up, ChatGPT also provided me with some extra tips that point out the subtle signs of an AI-generated scam call:
- Odd pacing or timing: Slight delays before answering, or responses that don’t quite match your question.
- Overly “perfect” or flat tone: Some AI voices sound too clean or emotionally off.
- Struggles with interruptions: If you cut them off or change topics suddenly, AI often falters or gives generic responses.
Protecting yourself in real-time and downloading apps for extra safety
ChatGPT also shared practical ways to protect yourself if you’re caught in the middle of an AI scam call.
Two tips stood out. First: ask a personal question that only the real person would know. Second: set up a family “safe word” — a phrase only your inner circle would recognize. Both make it much harder for a scammer using voice cloning to fake their way through a call.
For extra protection, ChatGPT also recommended using spam-filtering apps like Robokiller, Hiya and Truecaller to block suspicious calls altogether. It also suggested registering your number with the National Do Not Call Registry to cut down on telemarketing and spam.
Stay aware, stay protected
AI scams are getting smarter — and more convincing by the day. The best defense isn’t fear, it’s awareness. The more you know about how these scams work, the easier they are to spot before any damage is done. I’m making sure my own family knows what to look for — and it’s worth doing the same for yours.
