Fraudsters are reportedly trading massive scam operations for smaller and more profitable schemes using artificial intelligence (AI).
That’s according to Giles Thomson, president of the Financial Action Task Force (FATF), who outlined the fraud landscape in a Monday (Sept. 14) interview with the Financial Times.
As the report notes, fraudsters have for years used giant compounds housing hundreds of workers to pull off global scams. Thomson said that while authorities have become skilled in hunting down those operations, they are now seeing gangs AI.
In many cases, these scam compounds “are quite substantial buildings — you can see them on satellite imagery,” Thomson told the FT. “But we’re starting to see this spread to one or two people in a basement with a very big server.”
Criminals, many in Southeast Asia, have become adept at building relationships of trust with targets, before cheating them using cryptocurrency and investment frauds or romance scams. With AI, complex schemes become much simpler, letting scammers impersonate people with deepfakes or set up phony websites.
“We’re starting to see people using chatbots to go online and impersonate people and actually entice people into a relationship,” he said. “We’re seeing it in other areas of money laundering through deepfakes and clones to get onboarded by banks.”
The FT report cites data from Interpol showing that AI-enabled fraud was 4.5 times more profitable than traditional methods, with overall fraud-related reports climbing 54% since 2024.
“Agentic AI systems can autonomously plan and execute complete fraud campaigns — from reconnaissance to ransom demands,” the agency said.
As the FT notes, the FBI said earlier this year that it had gotten complaints about $893 million of losses from fraud carried out with the help of AI in 2025, saying the content in question is “increasingly difficult to detect and easier to make.”
PYMNTS wrote earlier this year about how artificial intelligence allows scammers to skirt traditional fraud detection, which had always relied on spotting outliers like abnormal transaction patterns or behavioral irregularities during onboarding.
“If you are sitting in a café or restaurant, people’s behavior typically is fairly similar,” James Mirfin, senior vice president, Global Head of Risk and Security Intelligence Solutions at Visa, told PYMNTS in March. “But you can spot someone that looks a bit nervous or twitchy. It’s the same in banking. Historically, bank tellers were looking for anomalous behavior.”
But now, AI systems trained on large datasets can now generate borrowers specially designed to evade those signals.
“If a human can do it, we are now at a stage where the machines can do it in plausible ways,” Adam Hiatt, vice president of fraud strategy at Spreedly, said. “It’s an arms race.”
