OnlyFake’s founder says he added generative AI to a service that made convincing fake identity documents, helping turn painstaking forgery into a product criminals could order at scale. 404 Media’s account traces how the operation grew, and why automated identity checks are part of the story.
Watch Desk analysis
What happened
Yuriy Nazarenko told 404 Media that he began by editing fake identity documents manually, then built a website to automate their production. The site initially automated the process without AI; later, he says, his team added AI-generated faces and backgrounds. Customers could provide personal details and photographs to create realistic images of documents, including pictures made to look as if an ID had been photographed on a surface.
404 Media says its account draws on interviews with Nazarenko and his lawyer, court records and the reporter’s own investigation. It reports that the US Department of Justice said OnlyFake generated at least 10,000 fake IDs and made $1.2 million. The article also describes the reporter testing generated IDs against an online exchange’s verification process.
Why it matters
The useful distinction is that AI was not the origin of the service’s automation, but it became part of a tool that made realistic fake documents easier to generate. That puts pressure on identity checks that rely on customers submitting document images, rather than establishing that a document or its holder is genuine by other means.
The story also shows how a small online operation could grow into a criminal service with subscriptions and bulk generation. A convincing document image is not proof of identity, however polished the pixels may be.
Our read
This is a concrete account of AI being added to an existing fraud operation, not a claim that AI alone invented identity fraud. The practical lesson is for organisations designing verification: a document upload needs scrutiny beyond whether the image looks plausible. The account is compelling because it describes the operation and a reported test, not because one story settles how every identity-check system performs.
What to watch
- Whether identity services disclose how they respond to realistic, AI-assisted document images.
- What further court records or law-enforcement actions show about the operation and its customers.
- Whether verification systems rely less on submitted images alone.
Discussion spark: If realistic fake IDs can be generated at scale, should online services rely less on document images, even if stronger checks make signing up harder for legitimate customers?
Sources and evidence
- Source update (5 November 2026, 14:16 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.