Social engineering of a person
Deepfake voice
Also called: deepfake video call
A deepfake voice is synthetic speech, made with AI, that imitates a real person closely enough to fool listeners. Attackers use cloned voices, and sometimes faked video, in calls and voice messages to pose as executives, officials or family members and persuade a target to share access, approve a payment or move money.
How it works
The attacker collects a few samples of the person's voice from videos, interviews or voicemail and uses AI tools to generate new speech. The fake message or call is paired with an urgent request and a reason to switch to a channel the attacker controls, which makes it harder for the target to check who is really there.
A real example
In May 2025 the FBI warned that malicious actors had sent text messages and AI-generated voice messages claiming to come from senior US officials, to build rapport before gaining access to personal accounts.
Source: Senior US Officials Impersonated in Malicious Messaging Campaign (I-051525-PSA) — FBI Internet Crime Complaint Center (IC3), 2025-05-15
The Identity Attack Ledger holds 12 cited incidents for social engineering of a person in California breach filings (January 1, 2023 – August 14, 2026), each quoted from its filed letter.
How to stop it
Treat a familiar voice as unverified. Confirm sensitive requests through a known contact method, and require identity checks for resets and payments that a voice alone cannot pass.
Related terms
Sources
- Senior US Officials Impersonated in Malicious Messaging Campaign (I-051525-PSA) — FBI Internet Crime Complaint Center (IC3), 2025-05-15
Last reviewed Oct 2, 2026