Dark web cybergangs are selling AI fraud tools as a service

If you thought dark web cybercriminals were just isolated hackers typing in dark basements, think again. The reality is looking uncomfortably close to a venture-backed tech startup with its own marketing department.

David Maimon, a criminologist at Georgia State University, recently uncovered a bizarre 58-second promotional clip intended for dark web underground markets. The video stars a man going by Sanchez, draped entirely in black and speaking behind a skull mask, with ominous music and an associate seemingly digging an open grave in the background.

Sanchez was essentially giving a customer service update to worried buyers who had not seen him online for weeks. During the pitch, he boasted that four years ago it was just him and a partner selling fake Chase bank accounts, but today they run an office staffed by 30 people.

Turning fraud into a subscription service

According to Haywood Talcove, CEO of LexisNexis Risk Solutions Government Group, the operation is tied to Mega Darknet Market, one of the biggest cybercriminal enterprises in the world. Having made fortunes defrauding emergency government relief programs during the pandemic, these global syndicates are now scaling their reach by arming other criminals with generative AI.

Instead of just executing scams themselves, these cartels run a full fraud-as-a-service model. They supply street-level scammers across Russia, Nigeria, China, and elsewhere with mule accounts alongside custom generative AI and deepfake software.

The real-world fallout extends far beyond stolen funds. Talcove pointed out that low-level criminals are using these accessible deepfake engines to fuel sextortion rings, taking everyday social media photos of young people, creating sexually explicit synthetic images, and blackmailing them. Recent FBI reports indicate that dozens of related suicides have already occurred this year, disproportionately impacting young boys aged 10 to 17.

Why it matters

This dark web clip matters because it proves that generative AI has commoditized cybercrime. You no longer need programming skills or machine learning knowledge to deploy sophisticated deception; you can simply buy the tools off the shelf from organized crime syndicates for a few hundred dollars.

For regular internet users and families, the threat surface just got radically wider. When convincing video and synthetic imagery can be generated in seconds from any public profile picture, standard safety intuition breaks down. Public social media presence now carries an inherent vulnerability that individual users cannot easily patch.

From an institutional perspective, the defensive perimeter is collapsing. Traditional policing is built to arrest the visible pawns at the bottom of the pyramid. But the offshore syndicates operating these AI-as-a-service pipelines sit far beyond domestic legal jurisdiction and can siphon billions from public coffers or private citizens before disappearing into thin air. Unless cybersecurity moves from reactive verification to active threat disruption, defenders will remain hopelessly outgunned.