How to Spot a Deepfake or AI Voice Scam

AI can make an emergency scam sound more personal by cloning a familiar voice or face. The technology is new, but the defense is old and reliable: stop, verify through another channel, and never let urgency replace confirmation.

Does this affect you?

Use this for phone calls, voicemails, video calls, or messages that seem to come from a family member, boss, coworker, celebrity, or trusted contact but ask for urgent money or secrecy.

Recognize the warning signs

The voice may be convincing, but the scam pattern usually stays familiar.

  1. Urgency and secrecy appear together, such as do this now and do not tell anyone.
  2. The caller asks for money through wire transfer, gift cards, cryptocurrency, Zelle, or another hard-to-reverse method.
  3. The voice sounds close but has odd pauses, flat emotion, strange phrasing, or wording the real person would not use.
  4. Video may show unnatural blinking, mismatched lighting, out-of-sync lips, overly smooth skin, or awkward movement.
  5. The caller avoids an unscripted back-and-forth conversation, talks over questions, or blames a bad connection.

Verify independently

This breaks almost every version of the scam.

  1. Hang up and call the person using a number already saved in your contacts.
  2. Ask a question only the real person would know that is not visible on social media or public records.
  3. Set a family safe word ahead of time for emergency money requests.
  4. Use another channel, such as texting the person or contacting another family member.
  5. Tell the caller you will verify and call back, then hang up even if they pressure you.

If someone was targeted or paid

Speed matters most after money moves.

  1. If no money was sent, stop contact and verify before doing anything else.
  2. If money was sent, contact the bank or payment service immediately to ask about reversal or recall.
  3. File a report with the FBI Internet Crime Complaint Center at ic3.gov.
  4. File a report at reportfraud.ftc.gov.
  5. Warn family members, especially older relatives, about the specific script used.

More control

Public voice clips can be enough

A few seconds of online audio can sometimes help generate a convincing clone. Public videos and voicemail greetings carry a small privacy tradeoff.

Not every emergency scam uses AI

Many still use generic voices and excuses like a cold or bad connection. Verify every unexpected emergency request the same way, whether it sounds real or not.

Sources

  • FTC Consumer Advice – Scammers use AI to enhance family emergency schemes (2025)
  • FBI IC3 – Public service announcement on AI-generated content in fraud schemes (2025)
  • CISA – Deepfakes and synthetic media (2025)
Disclosure: This post may contain affiliate links which means I may receive a commission for purchases made through links. I will only recommend products that I have personally used! Learn more on my Private Policy page.
A thoughtful woman reads a newspaper while enjoying coffee at an indoor workspace.

DEALWEEK

SUBSCRIBE AND GET 20% OFF YOUR NEXT ORDER! OFFER ENDS SOON - DON’T MISS OUT!

We don’t spam! Read our privacy policy for more info.

Shopping Cart