AI-generated voicemail is pre-recorded synthetic audio delivered through voicemail, voice notes, or asynchronous messaging. Unlike live attacks, the attacker records once and uses the same clip across many targets, lowering effort per attempt.

• Synthetic audio is generated using a cloned voice or generic TTS model
• The same clip is sent to thousands of recipients across phone, WhatsApp, Telegram, or email attachments
• Targets often respond by calling back, opening a follow-up live channel where additional fraud occurs

• "CEO" voicemails requesting urgent action from finance or HR
• Fake delivery notifications driving recipients to malicious callback numbers
• Family-emergency scams targeting consumers with cloned relative voices
• Coordinated extortion campaigns sending the same threat audio to many contacts
• Pre-recorded fake testimonials used to influence procurement or hiring decisions