AI-enabled scams surged in 2025 as crypto fraud became more industrialised, with scammers scaling operations using automation and businesslike efficiency, according to TRM Labs.
TRM Labs said the use of large language models in scams rose fivefold last year, helping fraudsters expand outreach, tailor messages across languages and make schemes more convincing at scale.
The firm estimated that $35 billion in cryptocurrency was sent to scammer addresses in 2025, down slightly from $38 billion a year earlier, even as overall illicit crypto activity climbed sharply.
“Large language models enable scams to cross language and cultural contexts with less friction, while AI-generated images, voice cloning, and deepfake videos reduce the cost of creating convincing personas,”
TRM Labs said in its 2026 crypto crime report.
TRM Labs said scammers are increasingly combining tactics, often starting with romance scams, shifting to fake investment offers and ending with advance-fee or tax-related fraud once trust is established.
“A defining trend in the current state of crypto fraud is the convergence of distinct scam typologies,”
The firm said, adding that layered deception makes fraud harder to detect and stop.
The report said illicit crypto wallets received about $158 billion in 2025, up roughly 146% from 2024, as sanctions evasion and improved detection tools revealed more activity despite illicit flows falling as a share of total crypto volume.