Thanks to the growing use of AI, and the ongoing industrialization of online fraud, crypto-based scams reached record levels in 2025, according to new data from research firm, Chainalysis.
The firm currently estimates that illicit addresses took in at least US$14 billion last year, and it projects that total will top US$17 billion as more illicit addresses are identified.
That represents a sharp increase from the US$9.9 billion in losses that it first reported for 2024, which rose to US$12 billion as data was finalized.
“The landscape has shifted dramatically as scammers industrialize operations, leverage AI tools, and increasingly target victims through sophisticated impersonation schemes,” the firm said in an initial report.
As a result of these developments, not only is the number of scams increasing, but the losses they generate are rising too.
Overall, the average scam loss rose by 253% year-over-year, to US$2,764 in 2025 from US$782 in 2024, the firm reported, “indicating more sophisticated targeting and victim profiling.”
In part, this reflects the growing use of AI tools to enable deepfakes, face swaps and other schemes that Chainalysis said “supercharged” the efficacy of crypto scams.
Using data on known scammers that acquired AI tools through blockchain-based payments, the firm estimated the scale and relative efficiency of AI-enabled scams — finding that schemes linked to AI tools extracted much more from victims on average, taking in an estimated US$3.2 million per operation, versus US$719,000 for more traditional operations.
The firm also reported that its analysis showed that AI-linked scams tended to scale more quickly and inflict more damage than scams not using AI.
“The increased transaction volume indicates that AI is enabling scammers to reach and manage more victims simultaneously, a trend consistent with the industrialization of fraud we’ve been tracking,” it said, adding that it appears that AI is also “making scams more persuasive.”
“We are moving toward a future in which virtually all scams will incorporate AI into their operations to some degree,” it said.
Additionally, the firm noted that the increase in crypto-related scams also reflects the development of fraud infrastructure at an industrial scale, such as firms offering “phishing-as-a-service” and selling “scam kits” for as little as $20.
The firm also reported that scams that involve impersonation of either government officials, or private sector players, “exploded” in 2025 — with the volume of these types of scams rising by 1,400% in the year, and the average loss to these schemes jumping 600%.