← Volver al blog
Amenazas IAAltaDestacado

Voice cloning: How a familiar voice can support an emergency scam

Recognising a voice does not establish identity. Verify the claimed emergency through an independent channel.

Revisión editorial:

El texto de este artículo está disponible en inglés en esta versión del sitio.

Contexto y fuentes

The FBI’s December 2024 warning describes generated audio impersonating loved ones in supposed emergencies or ransom demands. It recommends checking identity through an independently known contact route. A familiar voice can be one component of a deception, rather than reliable proof of who is calling. [1]

Alcance y límites

This is not proof of a universal three-second requirement or perfect imitation. An unusual voice alone does not establish that AI was used. [1]

Explicación sencilla

A hypothetical call sounds like a family member but creates intense time pressure. The critical check concerns the alleged situation, not just the sound of the voice.

Recomendaciones prácticas

We recommend agreeing in advance that suspicious emergency calls may be ended and independently checked. Keep trusted contact details available. A private family code word may help but should not replace independent contact. Do not send money to an unknown recipient under pressure. If immediate danger is genuinely suspected, contact the appropriate emergency service.

Fuente primaria / referencia

  1. [1] FBI/IC3 (03.12.2024): Generative AI and financial fraud