AI turns impersonation scams into a crypto cash machine

The spectacular and concerning growth in impersonation scams last year was also supported by new AI and phishing-as-a-service tools, which help criminals reach far more victims and steal significantly more money.

In total, crypto scams netted at least $14 billion in 2025, representing a 16% increase from 2024, according to blockchain analysis company Chainalysis. However, in the same year, impersonation scams jumped by 1,400%, while the average amount of payments made to these criminals increased by over 600%, the analysts said, without specifying exact figures.

Taking into account all scams, the average payment increased by 253% to $2,764 last year. Criminals are busy impersonating both governments and private organizations.

For example, a Chinese-speaking cybercriminal group known as “Darcula” and the “Smishing Triad” is suspected of being responsible for the “E-ZPass” phishing campaign, which targeted millions of Americans using the E-ZPass electronic road toll collection system in 2025.

Additionally, last December, a Brooklyn resident was indicted for impersonating a customer support representative at the Coinbase crypto exchange and stealing nearly $16 million.

Meanwhile, emerging AI tools are helping criminals make their operations more “profitable,” while evolving phishing kits also contribute to the industrialization of this type of crime.

According to Chainalysis, AI-enabled scams using face-swap software, deepfake technologies, and large language model solutions scale more quickly and hit their victims more severely.

“Our analysis reveals that, on average, scams with on-chain links to AI vendors extract $3.2 million per operation compared to $719,000 for those without an on-chain link [to Chinese AI vendors] – 4.5 times more revenue per scam,” the analysts said, adding that, with AI, median daily revenue increased more than 800%, to $4,838.

Additionally, the “professionalization and commercialization” of scammer tools help criminals scale their illicit operations. Data shows that scammers using phishing kits are 688 times more effective in dollar terms and four times more effective in average transaction size than regular scams. Also, scams that buy bulk social media accounts are 238 times more effective in dollar terms and twice as effective in average transaction value compared to regular scams, according to Chainalysis.

For example, in the E-ZPass case, scammers used software from Lighthouse, a Chinese-language vendor that offers cybercriminals “phishing for dummies,” with hundreds of templates for fake websites, domain-setup tools, and features designed to evade detection. In this group, different actors specialized in distinct parts of the scam and fraud supply chain, such as phishing software and templates, lists of potential victims, tools to send fraudulent text messages, monetization of stolen sensitive information, and online recruitment and collaboration forums.

While Chainalysis admits that there are no silver bullets to solve this problem, a stronger focus on fraud prevention and cross-border law enforcement cooperation might help slow this scam virus from spreading.

“However, we are also seeing a step change in law enforcement’s ability to respond,” Will Lyne, Head of Economic & Cybercrime at the Metropolitan Police, was quoted as saying in the report, adding that new capabilities and tools help law enforcement better identify criminal networks, seize illicit assets, and disrupt criminal activity.

Unlock more exclusive Cybernews content on YouTube.