Scammers used new tools to widen their reach and to seem more real. According to TRM Labs, the use of large language models in scams jumped fivefold in 2025, helping fraudsters write believable messages, run many conversations at once, and trick people in different languages.
AI Tools Helping Con Artists Build Trust
Reports say AI images, voice cloning, and deepfakes are cutting the cost of making fake people who look and sound legit. These tricks have fed a pattern where criminals first make a target feel safe and then ask for money.
In some cases, a romance angle is used to win trust, and that trust is later turned into fake investment offers or bogus tax demands. This staged approach has let scams run longer and capture bigger sums from fewer victims.
A Rise In Industrial-Scale Fraud
Behind many of these schemes are groups that act like small companies. They hire people, sell tools, and reuse scripts to run campaigns in many places.
Some providers now sell phishing kits or offer AI-as-a-service to automate messages and replies, lowering the bar for new fraudsters and making scams easier to copy and spread.
Deepfake Calls And Targeted Hacks
Reports note that attackers have even used fake video calls to trick crypto workers into installing malware. In several incidents, victims were invited to what looked like normal Zoom meetings, only to find AI-generated faces on the screen.
When the meeting “needed a patch,” victims were urged to install what was actually malicious software. These methods have been linked to North Korea–connected groups and were flagged by security researchers last year.
Crypto Price Action Enters The Story
While the scams became more sophisticated, the market evolved too. Bitcoin was trading in the range of $88,000 to $90,000 in late January 2026 as investors considered macro news and policy developments.
This market context is important: as prices increase, the urgency and authenticity of crypto scams may seem more plausible, and the risks for both victims and law enforcement may be higher.
Scam Proceeds Compared To Illicit Flows Overall
Illicit inflows to crypto assets reached a record high of $158 billion, a substantial increase due to improved monitoring that brought more illicit activity to light.
Meanwhile, scam-related wallets saw a slight decrease in proceeds to around $35 billion in 2025, from $38 billion in the previous year.
However, the total volume of criminal activity increased substantially, even as the portion attributed to scams increased marginally.
It appears that scam-detecting technology is improving, but scams are evolving rapidly. The increasing use of AI-based tools makes generic advice less helpful, as the scams now sound more authentic.
Featured image from Unsplash, chart from TradingView
