A new classification of criminal has quickly emerged. One that identity verification and fraud prevention firm Socure Inc. labels the AI-Fluent Fraudster.

In a new report, “From the Frontline: Fraud’s Napster Moment,” Socure explains why this new classification reflects the burgeoning impact of artificial intelligence. Not only is AI making it easier for organized groups to commit payments and financial fraud, everyday consumers are using it, too.

“On the consumer side, AI has removed the technical barrier to entry,” Mike Cook, Socure head of fraud insights, tells Digital Transactions News. “In our survey, 13.4% of respondents said they had used AI to commit at least one type of fraud, and more than half of those AI-fluent fraudsters showed no evidence of committing fraud without it. These aren’t career criminals picking up a new tool. They’re ordinary people, most commonly men under 30, driven by curiosity as much as by money.”

Socure found that 62.7% of AI-Fluent Fraudsters are male compared with 55.1% of those who commit traditional fraud. Incline Village, Nev.-based Socure gathered data from a survey of 1,250 consumers and data from its global identity and fraud intelligence network for the report.

Many of them are using AI to commit fraud—first-party fraud at 81% is the most common—seem to do so out of curiosity.

“Curiosity was their top motivation (41.8%), ahead of economic hardship (37.7%). Fraud has always been driven by need or greed, so curiosity is new,” Cook says. “They also feel removed from it: 16.5% said using AI made it feel like they weren’t the ones committing fraud, nearly three times the rate of traditional fraudsters (6.1%). First-party fraud has generally been driven by economic hardship and downturns in the economy. With curiosity taking the lead in motivation, I have concerns that even in a solid economy, first-party fraud rates may continue to grow at substantial volumes.”

These consumers are using AI to create fake hardship letters, fake income or bank statements, skipping a BNPL payment, faking a reason for a refund, and avoiding paying their credit card bills.

“First-party fraud is the easiest place to start because it can be rationalized as a way to recover money that someone believes they are owed. If the claim succeeds, the consumer learns what the system will accept. The same process can then be repeated across merchants, lenders, gaming platforms, and other industries,” the report notes.

Among organized criminals, AI is helping them run their operations more like a business where they can automate attacks and mass produce identities for repeatable profit, Cook says. “Ring members committing first-party, synthetic identity or identity theft were far more likely than solo actors to use AI (67.3% vs. 42.9%) and more than twice as likely to work across multiple fraud types.”

AI’s impact on the ease to commit fraud is worrisome. “What worries me is that first-party fraud looks like a gateway to other types of fraud. Synthetic fraud has been flat or slightly down over the past year. It used to be what everyone did. However, that could rise again as people learn they can use AI to build synthetic identities,” Cook says.

“That’s what makes it more important to look beyond the image. Synthetic documents and deepfake imagery are easy for fraudsters to produce, so it isn’t enough to ask whether a document is real. A friend of mine who likes to say, ‘If you’re solving for a deepfake, you’ve already lost,’ and I think that’s right. You have to look at the bigger picture, including device and location data.”