Deepfake fraud costs $3.7 billion worldwide, prompting new detection tools

A recent Surfshark study found that deepfake fraud has caused at least $3.7 billion in losses globally, with social‑media platforms responsible for 47 % ($1.73 illion) of the recorded damage. Impersonation scams accounted for $911 million, fake‑job candidate schemes $100 million, and other channels such as phone calls, video platforms and messaging apps contributed the remaining losses.

In response, several security firms have launched specialised detection products. Modulate offers voice‑native models that analyse audio spectrograms and micro‑temporal anomalies to flag deepfake audio. Resemble AI uses neural audio watermarking that embeds an imperceptible signal within the waveform to verify authenticity. Polygraf AI introduced Meeting Guard, a real‑time AI fraud detector for virtual meetings that checks voices against deepfake threats and protects against data leakage. These technologies aim to curb the rapid expansion of deepfake‑enabled scams across digital communications.