The application looks real and that is the problem. 

Synthetic identity fraud is one of the most damaging schemes in auto lending today — and one of the hardest to stop. A fabricated borrower with a carefully constructed credit history, a real Social Security number and a plausible employment profile can move through underwriting without a single red flag. 

The loss does not appear on Day One. It appears months later, when the borrower disappears. 

Synthetic identity fraud works precisely because credit models are designed to reward the behavior that makes it succeed. 

A fraud ring builds a synthetic identity slowly. It opens a secured card. It makes small payments. It establishes tradelines. Over time, the credit profile looks like a real person’s. The model sees a responsible borrower with an established history. Larger credit lines get approved, then auto loans. 

The model never had a chance — because what it was evaluating was never real. 

Why credit history is the wrong signal 

Credit risk models are built to predict repayment probability. They are not built to verify identity. 

A synthetic identity that has been cultivated over 18 months will score better than a real borrower with a thin file. The score reflects the behavior of the fraud scheme, not the creditworthiness of a person. By the time a lender realizes the profile was manufactured, the vehicle is gone and the loss is on the books. 

In fact, 75% of auto lenders reported that identified fraud increased over the prior 12 months, according to Informed.IQ‘s 2026 Auto Finance Fraud Intelligence Survey. The underlying driver, increasingly, is organized fraud activity — rings that operate across institutions, build multiple synthetic profiles simultaneously and hit multiple lenders in the same window. 

The scale is not accidental. 

Fraud rings run synthetic identity schemes as a business. The infrastructure exists to build profiles in volume, test them against lender systems and execute rapid bust-out sequences before any single institution can connect the pattern. 

Cross-lender blind spot 

No single lender sees the full picture. 

That is the structural vulnerability synthetic identity fraud exploits. A lender reviewing one application from one borrower with one set of documents has no visibility into what is happening at the institution three blocks away. The same synthetic identity can be used at multiple lenders simultaneously. 

The 2026 survey found that two-thirds of lenders rarely or never use historical document data for cross-application fraud checks. Only 12% reported relying heavily on an integrated data network for fraud detection. 

Organized fraud rings count on that gap. 

Synthetic profiles are often submitted with slight variations — a different employer, a marginally different income figure — to reduce the chance of an exact-match flag. The manipulation is invisible in single-application review. It becomes visible only when documents are cross-referenced across institutions and applications at scale. 

The document problem inside the identity problem 

Synthetic identity fraud does not operate in a vacuum. It arrives with supporting documents.

Generative AI has made it easier than ever to produce fabricated pay stubs and employment records that match the income profile of a synthetic identity. The credit history looks legitimate, the documents look legitimate and the deal gets funded.

The 2026 survey found that 80% of lenders have only slight or no confidence that their current systems can detect a gen AI-created counterfeit document. That confidence gap is the opening synthetic fraud rings need.

When a lender cannot reliably distinguish a real pay stub from a fabricated one, the document layer offers no protection. The application passes verification, the score passes underwriting and a borrower who does not exist receives a loan. 

Fraud is a network problem 

The lenders making progress against synthetic identity fraud are rethinking what verification means. 

Single-application document review catches single-application document fraud. Synthetic identity fraud is a network-level scheme and requires a network-level response. That means cross-referencing submitted documents against a consortium of previously processed records, not just validating what is in front of a reviewer today. 

Purpose-built AI solutions connected to data consortiums can surface the patterns that make synthetic fraud visible — profile reuse, document reuse, employer clustering, application timing anomalies. These signals do not exist in any one lender’s data. They exist in the aggregate.

A clean score does not mean a real borrower, and a funded loan to a synthetic identity is not a loan that is ever going to perform.

Synthetic identity fraud is a solvable problem, but solving it requires treating identity verification as a separate risk layer from creditworthiness — one that is evaluated before the application ever reaches underwriting, and one that draws on intelligence no single institution can build alone. 

Jessica Gonzalez is the vice president of customer success at Informed.IQ and has more than 15 years’ experience in the financial services industry, including tenures at Santander Consumer USA and Visa. 

Content sponsored by Informed.IQ