Illicit cryptocurrency flows reached a record $158 billion in 2025 as scammers armed with artificial intelligence tools expanded their operations globally, according to blockchain analytics firm TRM Labs, while the use of large language models in fraud schemes increased fivefold during the year.
What Happened: AI-Powered Scam Operations
TRM Labs reported that criminals deployed AI-generated images, voice cloning and deepfake technology to create convincing fake identities at lower costs than ever before.
The tools enabled fraudsters to write believable messages in multiple languages and manage numerous conversations simultaneously. Many schemes followed a pattern where criminals first established trust with victims before pivoting to fake investment offers or fraudulent tax demands.
Some attackers used fabricated video calls to trick cryptocurrency workers into installing malware.
Victims were invited to what appeared to be routine Zoom meetings, only to encounter AI-generated faces on screen and prompts to install malicious software disguised as patches.
Security researchers linked several incidents to groups connected to North Korea. Behind many operations were organized groups functioning like small companies, hiring staff, selling tools and reusing scripts across multiple campaigns.
Also Read: Why Central Banks Are Stockpiling Gold Instead Of U.S. Debt For First Time Since 1996
Why It Matters: Evolving Criminal Landscape
The $158 billion figure represented a substantial increase from prior years, partly due to improved monitoring that uncovered previously hidden activity.
Scam-related wallets saw proceeds decline slightly to around $35 billion in 2025 from $38 billion the previous year. However, overall criminal volume increased substantially even as the scam portion grew marginally.
Bitcoin (BTC) traded between $88,000 and $90,000 in late January 2026, a market context that may lend urgency and plausibility to cryptocurrency fraud attempts. Detection technology continues to advance, but AI-based tools have made generic prevention advice less effective as scams now sound more authentic.

