Key figure: Deloitte projects US generative-AI-enabled fraud losses to reach $40 billion by 2027, up from $12.3 billion in 2023 (Deloitte Center for Financial Services).
A deepfake voice scam clones a loved one's voice with AI to fake emergencies. Cybersecurity news flags it as a top family threat in 2026. This free kit gives your family a code-word plan.
Cybersecurity news in 2026 reports a nightmare: a parent gets a call sounding exactly like their child begging for help — but it is an AI-cloned voice. Attackers need only seconds of social-media audio to build the clone. The defense is low-tech: agree on a family code word no AI can guess, and verify urgent money requests through a second channel. This kit sets your family up in 10 minutes.
A deepfake voice scam uses AI to clone someone's voice — often a family member — then calls targets with urgent fake emergencies to steal money. Seconds of social media audio is enough to build a convincing clone.
It sounds like your grandson. It sounds like your daughter. It is not. These are deepfake voice scams — AI-cloned voices used to fake family emergencies and demand money. The U.S. Federal Trade Commission warns that a cloned voice can sound exactly like the real person, and its advice is blunt: do not trust the voice. Source: FTC consumer alert on AI-enhanced family emergency scams.
A code word is the one defense a voice clone cannot beat: a secret decided in person, asked on any panicked call. Generate one now, then tell every family member face-to-face. Never text it. Never email it.
Two messages per round. One follows a real reported scam script (EXAMPLE scenarios, illustrative only); the other is an ordinary call. Listen to both — buttons speak with your device's synthetic voice — then pick which one is the scam. You cannot rely on the voice. Listen for the script: panic, hurry, secrecy, strange payment.
Your downloadable one-page plan folds the code word, the protocol, and the FTC's core advice into a single page. Print it. Stick it on the fridge. The safest household is the one that already decided what to do.
The scam the FTC describes works because love is faster than doubt: a panicked voice that sounds like your grandson, an arrest or a wreck, money needed right now, and — critically — do not tell anyone else. Every element is engineered to stop you verifying. A code word agreed in advance breaks the spell in one question, because no cloning tool can guess a secret it has never heard. It is analog, it is free, and it is the one check that scales from children to grandparents.
Scammers clone a loved one's voice from a short audio clip, then call pretending there is an emergency — an accident, an arrest, a kidnapping — and pressure the victim to send money or share sensitive information. The FTC warns the clone can sound exactly like the real person, so the voice alone is never proof of identity.
Yes. A short clip from TikTok, Facebook, Instagram, or YouTube — or even a voicemail greeting — can be enough material for a convincing voice clone. Limit how much of your voice is public, set profiles to private, and treat any public audio as potential cloning material.
Create a family verification plan before a scam happens: agree on a private family code word decided in person and never shared by text or email, never send money based on one incoming call alone, and always hang up and call back on a saved number. Use the generator and test above to practice.
Not reliably. There may be small tells — odd pauses or unnatural phrasing — but AI voices are now close enough that listening is not a safe test. The defense is verification: hang up, call back on a saved number, and ask for your family's code word.
Secrecy plus urgency plus a hard-to-reverse payment (wire, crypto, gift cards) is the classic scam pattern. Do not send anything on that call. Hang up, verify the story through a separate channel, and report the attempt to the FTC at ReportFraud.ftc.gov.
Hang up immediately — even if the voice sounds exactly like your loved one — then call them back on a number you already have saved. Ask for your family's pre-arranged code word; scammers can't guess it. Never send money during that first call, and report the attempt to the FTC at ReportFraud.ftc.gov.
How do AI voice cloning scams work?
Scammers clone a loved one's voice from a short audio clip, then call pretending there is an emergency — an accident, an arrest, a kidnapping — and pressure the victim to send money or share sensitive information. The FTC warns the clone can sound exactly like the real person, so the voice alone is never proof of identity.
Can scammers clone your voice from social media?
Yes. A short clip from TikTok, Facebook, Instagram, or YouTube — or even a voicemail greeting — can be enough material for a convincing voice clone. Limit how much of your voice is public, set profiles to private, and treat any public audio as potential cloning material.
What is the best way to protect your family from AI voice scams?
Create a family verification plan before a scam happens: agree on a private family code word decided in person and never shared by text or email, never send money based on one incoming call alone, and always hang up and call back on a saved number. Use the generator and test above to practice.
Can you detect a deepfake phone call by listening?
Not reliably. There may be small tells — odd pauses or unnatural phrasing — but AI voices are now close enough that listening is not a safe test. The defense is verification: hang up, call back on a saved number, and ask for your family's code word.
What should you do if you receive an AI voice cloning scam call?
Secrecy plus urgency plus a hard-to-reverse payment (wire, crypto, gift cards) is the classic scam pattern. Do not send anything on that call. Hang up, verify the story through a separate channel, and report the attempt to the FTC at ReportFraud.ftc.gov.