Increasingly sophisticated AI allows scammers to replicate someone’s voice with an audio sample that could be as short as just a few sentences.

Voice impersonation scams are on the rise in the US. The advancement of artificial intelligence (AI) has made it easier for bad actors to mimic voices, tricking people into believing that their loved ones are in distress.

A person impersonating Ruth Card’s grandson called her to say he was in jail and needed cash for bail. She withdrew CAD3,000 ($2,207) from her bank before realizing the man was not her grandson. In 2022, impersonation scams were the second most popular scam in America, with over 36,000 reports of people being duped by those pretending to be friends or family. Over 5,100 of those incidents happened over the phone, accounting for over $11 million in losses. Elderly, easy to panic victims appear to be a choice target. (The Washington Post)

Voice impersonation

AI voice-generating software analyses a person’s voice unique features and searches a vast database of voices to find similar ones and predict patterns. The software then re-creates the pitch, timber and individual sounds of a person’s voice to create an overall effect that is similar. All it needs is a short sample of audio, taken from places such as YouTube, podcasts, commercials, TikTok, Instagram or Facebook videos.

A criminal impersonator can then make this “voice replica” speak out whatever they type into a text box.

Companies such as ElevenLabs, an AI voice synthesizing start-up founded in 2022, transform a short vocal sample into a synthetically generated voice through a text-to-speech tool. This software can be availed for free, or on subscription ranging between $5 and $330 per month.

ElevenLabs recently said it is developing checks such as not allowing free users to create custom voices as well as tools that can detect AI-generated audio.

Related Story:  Tech Giant Microsoft’s VALL-E AI Tool Can Reproduce Your Voice

Photo by Soundtrap on Unsplash