The Internet Crime Complaint Center's annual report for calendar year 2025, marking "a quarter-century of cybercrime reporting." Retrieved through the Internet Archive; ic3.gov blocks direct retrieval.
Headline figures
| Measure | 2025 |
|---|---|
| Complaints | 1,008,597 |
| Reported losses | $20.877 billion (+26% on 2024) |
| Average loss | $20,699 |
| AI-related complaints | 22,364 |
| AI-related losses | $893,346,472 |
| Cryptocurrency-related complaints | 181,565 ($11.37 billion) |
Losses concentrate sharply by age: complainants 60 and over accounted for 201,266 complaints and $7.748 billion, a 37% rise in complaints and 59% rise in losses over 2024, with an average loss of $38,500 and 12,444 complainants losing more than $100,000.
AI as a descriptor, not a crime type
The report's methodological choice matters for reading the figures: "AI Related" is a descriptor applied on top of a selected crime type, defined in the appendix as meaning "information reported contains a reference to artificial intelligence (AI)." It records what complainants mentioned, not what forensic analysis established.
The report frames AI neutrally — "like most technology, artificial intelligence (AI) is a tool which can be used for legitimate, helpful purposes or for criminal motives" — and locates the change in scale and quality rather than in kind: "People have manipulated video and audio similarly for decades, but the widespread availability of this developing technology makes it possible to create high-quality content" in "mass quantities," with synthetic content "becoming increasingly difficult to detect and easier to make."
Where the AI losses fall
AI-related losses are dominated by a single category. Of $893 million total:
| Crime type | AI-related loss | AI-related complaints |
|---|---|---|
| Investment | $632,041,188 | 4,356 |
| BEC | $30,256,592 | 135 |
| Tech/Customer Support | $19,457,078 | 574 |
| Confidence/Romance | $19,041,653 | 626 |
| Personal Data Breach | $18,767,964 | 1,204 |
| Employment | $12,550,185 | 691 |
| Extortion | $2,940,642 | 1,764 |
The complaint-count ranking differs from the loss ranking: extortion generates the second-highest AI-related complaint count (1,764) but the twelfth-highest losses, while BEC produces the second-highest losses from only 135 complaints.
The report's most consequential caveat concerns undercounting: "overall losses to Investment scams exceeded $8 billion, demonstrating that many victims do not realize the extent AI may be involved in scams." Since the descriptor depends on complainant awareness, the $893 million figure is a floor on AI-involved losses rather than an estimate of them.
Mechanisms described
- BEC. "Chat generators can quickly create official-sounding emails mimicking a company's CEO or other officials," with voice cloning used "to request wire payment." The report notes explicitly that "there are multiple BEC tactics, and not all are AI-enabled."
- Confidence/romance. Fake profiles and AI-generated scripts "to make speech more believable," including distress or grandparent scams "in which voice cloning technology is used to mimic the sound of a loved one in distress" — over $5 million in 2025, and the report notes the scam "is evolving to mimic other family members or close friends in different types of emergency scenarios."
- Employment. Voice spoofing or deepfakes during online interviews, where "the actions and lip movement of the person interviewed on-camera do not completely coordinate with the audio," and "actions such as coughing, sneezing, or other auditory actions are not aligned with what is presented visually." The report notes the motive differs from other categories: "there does not seem to be significant dollar loss associated as the goal generally appears to be gaining access to private computer networks."
- Investment. AI used to "quickly generate thousands of conversations that appear different to each prospective victim," with investment clubs employing "AI-generated videos and voices of celebrities, CEOs, or trusted figures."
The scaling argument is common to the BEC and investment mechanisms: what AI supplies is per-target personalization at volume, which defeats the duplicate-message heuristics that previously made mass fraud detectable.
Relationships
- supports: AI and Cybersecurity — the federal data point on AI-enabled consumer fraud
- related: AI Voice Cloning — voice cloning appears across BEC, distress, and employment scams
- related: Synthetic Media / Deepfakes, Financial Services — AI Deployment, TAKE IT DOWN Act