Deepfake fraud has moved from rare attack to repeatable crime. Verified reports from Onfido, Sumsub, Europol, and UK Finance show sharp increases through 2024, and those trends frame what 2026 data will likely confirm. Global losses now exceed $25 billion a year. If you were targeted, start with what to do after an AI scam.
These growth rates matter because they outpace most identity fraud categories. A 3,000% rise in detection means criminals are scaling deepfake use faster than many verification systems can adapt. The EU complaint surge also shows victim reporting is increasing, not necessarily that the underlying crime is new. Implications: first, deepfake fraud is becoming a volume attack, not a rare event. Second, identity verification alone cannot stop voice cloning. Third, victims should document suspicious calls early. If you received a suspicious AI call, review voice cloning scam recovery.
The Hong Kong case shows how a single deepfake meeting can bypass multiple layers of human verification. The $25 million loss was approved after employees saw and heard what they believed were company executives on a video call. It was not a technical exploit alone; it was a social engineering attack. Victims often need both financial recovery and emotional support. Report losses to FTC Report Fraud and FBI IC3. See how to get money back after an AI scam for recovery steps.
Deepfake Fraud Growth Indicators
| Stat | Detail | Source |
|---|---|---|
| 3,000% increase | Deepfake fraud attempts detected on Onfido’s platform rose 3,000% in 2023 compared with 2022. | Onfido, 2024 |
| 10x increase | Deepfakes detected on Sumsub’s platform increased tenfold between 2022 and 2023. | Sumsub, 2023 |
| 2,137% increase | AI-generated synthetic identity fraud grew 2,137% from 2021 to 2023. | Sumsub, 2023 |
| 240% increase | Deepfake fraud complaints across EU member states rose 240% between 2022 and 2023. | Europol, 2024 |
This split can mislead. The 4% share that is fraud still produced multi-billion-dollar losses because fraud cases are monetized at far higher rates than most content. A single impersonation video can move millions. Non-consensual content causes severe harm and often falls under different legal and support pathways. If you appear in a manipulated video, see deepfake video scam what to do.
Financial Losses and Landmark Deepfake Fraud Cases
| Stat | Detail | Source |
|---|---|---|
| $25 billion annually | Estimated global financial losses from deepfake-enabled fraud exceeded $25 billion per year as of 2024. | Sumsub, 2024 |
| $25 million single attack | A deepfake business email compromise on a Hong Kong multinational cost $25 million (HKD 200 million) in January 2024. | Hong Kong Police Force, 2024 |
| EUR 220,000 voice scam | A UK energy company CEO was tricked by an AI-generated voice call into a fraudulent wire transfer in 2019. | Wall Street Journal, 2019 |
| $1.7 billion investment scams | Deepfake-powered investment scams using fake celebrity endorsements generated an estimated $1.7 billion in victim losses globally in 2023 across at least 37 countries. | Global Anti-Scam Alliance, 2024 |
| GBP 485 million impersonation losses | UK fraud losses from impersonation, including AI voice and video deepfakes, totaled GBP 485 million in 2023. | UK Finance, 2024 |
Reported Deepfake Content Split
| Stat | Detail | Source |
|---|---|---|
| 96% non-consensual sexual images | Of deepfake videos online in 2023, 96% were non-consensual sexual images. | Sensity AI, 2023 |
| 4% fraud, disinformation, or impersonation | The remaining 4% were financial fraud, disinformation, or impersonation. | Sensity AI, 2023 |
Frequently Asked Questions
What is the latest full-year deepfake fraud total for 2026?
No public agency has released a complete 2026 full-year figure yet. The most recent verified annual benchmarks are from 2023 and 2024. Treat any precise 2026 total with caution.
How much money do deepfake scams steal annually?
Verified estimates put global deepfake-enabled fraud losses above $25 billion a year as of 2024. Single incidents range from EUR 220,000 to $25 million. Fake celebrity investment scams alone caused about $1.7 billion in victim losses in 2023.
Are most deepfake videos used for financial fraud?
No. Sensity AI found that 96% of deepfake videos online in 2023 were non-consensual sexual images. Only 4% were financial fraud, disinformation, or impersonation. That small share still drives the bulk of financial losses.
Which type of deepfake fraud causes the largest individual losses?
Business email compromise and fake executive video calls show the largest verified single losses. A Hong Kong company lost $25 million in one attack. AI voice impersonation is also common and feeds broader impersonation fraud.
What should I do if I am a victim of a deepfake scam?
Stop contact, preserve screenshots and recordings, and notify your bank or payment provider. File a report through how to report an AI scam. Speed is the strongest factor in recovering funds.
