Safe word: What is it and why do you need one?

AI voice scams are no longer as niche as they once were. The technology is both cheap and convincing: a worrying combination that has lowered the barrier to entry for would-be scammers. According to a Hiya report from March, one-in-four Americans say they’ve received a deepfake voice call in the past 12 months. A further quarter (24%) say they can’t tell the difference.

The same study reveals that twice as many Americans think scammers are winning the battle with mobile carriers over voice clones. But it’s not a foregone conclusion. The technology might be impressive, but a simple technique could stop the scammers in their tracks: a pre-agreed safe word.

We all think we could recognize a loved one’s voice. But cloning technology is advancing at a dizzying rate. All scammers need is a few seconds of audio to create a convincing deepfake. And that can be scraped from our social media posts, if we haven’t restricted access via the privacy settings. Or possibly work-related content that ends up being posted online.

So once they’ve cloned a voice, what do they do with it?

One of the most prominent scams involves cold calling a relative and playing deepfake audio designed to make them think their loved one has been kidnapped. These virtual kidnapping scams may be made to appear more realistic by citing personal details also harvested from the same social accounts. Scammers may also monitor the victim’s social media profile, or those of their relatives, in order to choose the best time to strike. It can be more useful for them if the victim is abroad at the time and therefore not checking their phone for incoming calls.

The key for the fraudsters is to keep the victim’s relatives guessing, and traumatized. To this end they’ll use the cloned voice only for a few seconds at a time, mix it with sobbing noises, and potentially introduce background noise to make the whole thing sound more plausible, and chaotic.