Synthetic Voice Vishing

Synthetic voice vishing (a combination of "voice" and "phishing") uses AI voice-cloning technology to generate audio that closely mimics a specific real person's voice, often trained on just a few seconds of publicly available audio such as a social media video or voicemail greeting. The cloned voice is then used in a phone call to impersonate that person convincingly, most commonly a family member claiming to be in an emergency and needing money urgently, or a company executive instructing an employee to make an urgent wire transfer.

The emotional pressure and apparent authenticity of hearing a genuinely familiar voice, rather than a stranger reading a script, is what distinguishes this from older impersonation scams and makes victims far less likely to question the request in the moment. Corporate variants have used cloned executive voices, sometimes combined with real-time video deepfakes on conference calls, to authorize large fraudulent fund transfers by convincing finance staff a legitimate internal request is underway.

Establishing a family safe word used to verify identity during any emergency call, and independently calling back a known number for the supposed caller before acting on an urgent financial request, are the most effective defenses, since they do not rely on being able to detect the clone by ear alone.