In 2025, global damage from fraud using deepfake technology reached a record $1.1 billion, three times higher than the 2024 figure of $360 million. According to a Surfshark study, social networks accounted for 83% of all financial losses, making them the main source of fraudulent traffic.
The research methodology is based on an analysis of data from the AI Incident Database and Resemble.AI for the period from 2017 to September 2025. The report includes incidents involving AI video, audio, and images that have been covered in the media. Analysts classified the cases into 14 categories of fraud and identified the platforms on which they were distributed. In cases where several social networks were mentioned, the financial damage was distributed equally among them.
Three platforms generate 93% of all monetary losses recorded on social networks. Facebook leads the trio with $491 million, followed by WhatsApp with $199 million and Telegram with $167 million.
Investment schemes were the main driver of growth, accounting for 80% of the total. Fraudsters create videos featuring businesspeople, experts, and celebrities, encouraging people to invest money in non-existent projects. The damage amounted to $886 million.
The second trend was romantic fraud. Using AI to imitate the appearance and voice of their ‘partners,’ criminals swindled their victims out of approximately $10 million. In 57% of cases, women were the targets of such attacks.
- AI Incident Database: https://incidentdatabase.ai/
- Resemble.AI: https://www.resemble.ai/deepfake-database/
- Table of materials: https://docs.google.com/spreadsheets/d/1y2-xlH0z5oyRVU0Lx9F0CQUkYBbyS7YsbWwbTwgSVdE/edit?gid=906959336#gid=906959336
