Egidio — Threat Lab · AI & Deepfakes Report · v1.0 · July 2026 · License CC BY 4.0 · egidio.app/en/laboratoire/generative-ai-voice-deepfakes/
A documented change in scale
In 2026, the FBI introduced a specific "AI-related" descriptor into its IC3 complaint center — an official acknowledgment that these scams now form a category of their own. In the very first reporting cycle under this classification: more than 22,000 complaints, for close to $900 million in losses.
A documented case
In early 2024, an employee at a multinational company was deceived by a video conference entirely composed of AI-generated participants, impersonating several company executives — to the point of authorizing a $25.6 million wire transfer. The case was documented by Hong Kong police and widely covered by specialized press.
Why it's different from a classic phone scam
A fake-bank-advisor or fake-tech-support scam relies on a script and the scammer's conviction. A scam backed by generative AI can additionally reproduce a recognizable voice — a relative's, a boss's, a child's — from a very short audio sample, which makes the usual verification reflex ("I recognize their voice") far less reliable than before.
Anti-hype, by design. This report preserves the nuance of the source agencies: AI "amplifies existing techniques," it doesn't create an autonomous threat out of thin air. No catastrophic scenario is added beyond what the sources actually document.
How to cite this publication
Egidio, Threat Lab, "AI & Deepfakes Report," v1.0, 2026, egidio.app/en/publications/ai-deepfakes-report/
Rigor note. This report derives from the page Generative AI & voice deepfakes. No new figures are introduced here. License CC BY 4.0.
Learn more
Egidio — The Threat Laboratory, "AI & Deepfakes Report — Next-Generation Voice Fraud", egidio.app/en/publications/ai-deepfakes-report/. Licensed CC BY 4.0.
Free to reuse, including commercially, with attribution. Reuse terms.