Actually Works · Episode 38

01 / 38

Just ten seconds of your voice is all AI needs to clone it now — and scammers are already using it.

Scammers only need a few seconds of your real voice — from a public video, a social post, or a voicemail greeting — to clone it with current AI tools, then call someone you love pretending to be you, asking for money. The real defense isn't a setting in an app: it's a shared safe word only your loved ones know.

The exact clicks

  1. Get the people who'd ever need to reach you in an emergency together — in person, or a group chat everyone's already in. This doesn't work if only one person knows it.
  2. Agree on two random, unrelated words nobody could guess or find online — not a pet's name, not a birthday, not an inside joke you've ever posted. Example: a made-up pairing like 'lighthouse-pretzel'.
  3. Say the words out loud together once, so everyone has actually heard them spoken, not just read them in a text.
  4. Save it somewhere each person can find later without asking out loud in the moment — a private note on your phone, not just memory.
  5. Agree on the rule: if anyone calls sounding distressed and asking for money, gift cards, or crypto, the other person asks for the safe word before doing anything else.
  6. Agree on the fallback: if the caller can't give it, hang up and call that person back on the number already saved in your phone — never the number that just called, since caller ID can be faked.
  7. Test it once, on purpose, in a calm moment — not during a real scare — so everyone actually knows the drill when it matters.

What changes

  • Verified live (17.9.2026) against multiple independent 2026 cybersecurity/consumer-safety and voice-cloning-industry sources, cross-referenced rather than relying on a single outlet: current instant-cloning APIs (e.g. Gradium, Cartesia) specify a 10-second minimum sample; the caller-ID-spoofing mechanic and the safe-word defense are also current and consistent across sources.
  • A plain-language, personal-safety topic — a deliberate exception to this channel's usual demand-report-driven topic selection, made explicitly per the 16.9.2026 reach-first content decision (see channel/content-memory.md), same reasoning as episodes 36 and 37.

What it will not do

  • A safe word stops the specific 'fake emergency call' scam pattern described here — it doesn't protect against every kind of scam or fraud.
  • This only works if it's actually agreed upon in advance and everyone involved remembers it exists — it protects nothing if only one person set it up.
  • This episode covers the voice-cloning phone-scam pattern specifically, not broader AI-deepfake risks (video, images) which use different mechanisms and defenses.

Get the next one

One AI setup a week, straight to your inbox.

Also buildingAgently— the catalog for AI agents