Recorded global losses from deepfake fraud have reached $3.7 billion, according to research published by Surfshark on July 14, 2026. The study combined records from the AI Incident Database, Resemble.AI and the OECD covering January 2020 through June 2026, counting only cases verified by media reports with documented financial losses.

Losses totaled $83 million between 2020 and 2023. The figure rose to $335 million in 2024, reached $2.5 billion in 2025, and stood at $764 million for the first half of 2026. The 2025 and 2026 totals together account for 89 percent of all recorded deepfake fraud losses.

Social media platforms are the largest single origin, at $1.73 billion, or 47 percent of the global total. Much of that sum traces to criminals using synthetic video of celebrities and public figures to promote fraudulent investment schemes. Impersonation fraud ranks second at $911 million, a category covering synthetic facial verification videos, face swaps, voice cloning, altered identity documents, fraudulent account openings and loan application fraud.

Crypto ATM fraud produced $333 million in losses during 2025, with scammers posing as trusted contacts to pressure victims, often older adults, into sending funds through cryptocurrency machines. Fake job candidate schemes added $100 million, driven by AI generated resumes, face swapping and deepfake video interviews aimed at hiring processes for technology roles in the United States.

Everyday communication channels carried smaller totals. Phone calls accounted for $71 million in losses, video platforms and conferencing tools for $62 million, messaging apps for $41 million and email for $12 million.

Source: Surfshark - https://surfshark.com/research/chart/deepfake-fraud-origins