A $15 Service Used AI to Forge Identity Documents and Slip Past Crypto Exchange Checks
What happened
Early in 2024, a service called OnlyFake began advertising something that should not have worked: AI-generated identity documents convincing enough to pass the automated Know Your Customer checks that crypto exchanges use to comply with anti-money-laundering regulations. For fifteen dollars, a customer could order a fake driver's license or passport rendered by neural networks, choose from documents representing 26 countries including the United States, Canada, Australia, the United Kingdom, and multiple EU member states, and receive a finished image within minutes.
KYC verification systems at crypto exchanges typically ask new users to upload a photo of a government-issued ID, sometimes paired with a selfie. The checks compare document layout, fonts, and security features against known templates and flag images that look obviously manipulated. OnlyFake's output was designed to pass exactly those template checks. The service claimed its neural network process produced documents that looked photographically real rather than digitally constructed, which meant the tells that automated systems had been trained to spot were absent.
404 Media tested the claim and confirmed it. Reporters used an OnlyFake-generated image of a British passport to create an account on OKX, one of the larger global crypto exchanges, and successfully completed the KYC step. The test was published in February 2024 and named OKX specifically, though the service's own marketing suggested it had been used across multiple platforms over the preceding weeks.
OnlyFake went offline shortly after 404 Media published its investigation. By March 2024, a version of the service had relaunched with added disclaimers and new tools, including the ability to generate matching handwritten signatures alongside the forged documents. The relaunch framed the product as intended for entertainment or testing purposes, but the added signature capability made it more functional for fraud, not less.
The incident exposes a structural gap in automated identity verification: KYC systems reach a pass-or-fail decision from a document image, but nothing in most implementations creates a provable record of what a system did at each step, which features it evaluated, and why it accepted a given submission. When a forged document passes, there is no audit trail tracing the failure to a specific algorithmic decision or a specific visual feature the system was relying on. That absence means each successful bypass goes unexamined, and the next service can iterate on the same method without any institutional learning on the other side.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.