Home/Guides/Fake Payslip Detection for Loans
Lending
Fake payslip detection for the most-edited document in a loan application.
Income is the primary underwriting variable, and payslips are easy to fake. TamperCheck catches the three patterns that dominate loan fraud: net pay inflation on a genuine payslip, fully fabricated payslips from templates, and AI-generated payslips from text prompts. All three pass visual review. Forensic analysis catches what arithmetic and eyeballs can't.
- No document storage
- Digital, scanned or photographed
- ~1 minute per verdict
Last updated:
The problem
Why payslips are the go-to fraud target in lending
Payslips are the most-edited document in loan applications. Income is the primary underwriting variable, and a payslip is easy to fake: most look similar, the structure is well known, and the figures are small enough to feel low-stakes to fraudsters. The result is that payslip fraud accounts for the largest share of document fraud across personal lending, mortgages, and auto finance.
Ready-made payslip templates imitating real employers are sold openly online, and AI tools generate a convincing payslip from a text prompt. Payslips are also rarely verified at source - they arrive as "provided by the applicant", often with no independent check.
Three patterns dominate: net pay inflation (changing a single number), fabricated payslips (generated entirely from a template), and AI-generated payslips (created by a generative model from a text prompt). All three pass visual review. Even bank-statement cross-checks are unreliable - fraudsters know to inflate both documents consistently.
Fake payslip detection has to be forensic, not arithmetic. The signals are in the PDF structure, the font metrics, the producer metadata, and the spectral properties of any embedded images. TamperCheck inspects all of them automatically.
What we detect
Forensic signals that expose fraud
Net-pay arithmetic verification
Gross pay, deductions, tax, super/401k, and net pay must all reconcile. A single edited figure breaks the chain; TamperCheck catches it.
Font metric consistency
Edited fields often use slightly different font weights or spacing. Cross-field font metric comparison reveals these edits.
Producer metadata audit
Real payroll systems (ADP, Xero, Gusto, MYOB, BambooHR) produce PDFs with consistent producer signatures. Edits introduce new tool fingerprints.
AI-generated payslip detection
Fully synthetic payslips from generative AI tools have spectral signatures that real payroll-system PDFs don't. The CV layer is trained on these.
Employer branding and template fingerprints
Low-resolution logos, wrong company addresses, tax or company numbers that don't match the stated employer, and the structural fingerprints of known fake-payslip templates.
Year-to-date plausibility
YTD figures, pay period sequencing, and date consistency are cross-checked against the claimed pay frequency.
Bank statement cross-validation (optional)
When the borrower submits a bank statement, TamperCheck cross-validates that the net pay actually credits the account on the claimed date.
The solution
Payslip fraud detection that sits inside your loan origination workflow
Most lenders wire it into document collection. Every payslip an applicant uploads goes through the API. The verdict gates the application's progress to underwriting.
How it works
- Integrates with any LOS - Encompass, Calyx, nCino, MeridianLink, custom
- Runs on every payslip - not just flagged applications
- per-document pricing - a fraction of any approved fraudulent loan
- Verdict in about a minute - keeps origination workflows moving
- Risk score for auto-routing - clean payslips skip manual review
- Works with multiple payslips per applicant - cross-validates across them
- Zero storage - critical for borrower data compliance
Why us
What you won't get elsewhere
Findings, not a score
We name the field that changed and where it sits on the page.
Others hand you a number to justify.
One endpoint, every type
100+ document types, 190+ countries, no per-type setup.
Elsewhere, each new form is a project.
Catches what was never real
A dedicated pass for AI-generated and deepfake documents.
Template matching only catches edits.
Nothing is retained
Analysed in memory, then discarded. Never trained on.
Most services keep documents for QA.
Live in minutes
Published pricing, no minimum. Test it before you talk to us.
Platforms want a contract first.
Sits alongside your stack
Keep Onfido or Jumio for face match. We add the forensics.
Not a rip-and-replace.
FAQ
Common questions
How do you detect a fake payslip on a loan application?
Real detection requires forensic analysis, not just arithmetic. PDF structure, metadata, font metrics across every field, pixel-level signals on embedded images, the producer signature from the payroll system, and AI-generation spectral signatures. TamperCheck runs all of these automatically.
What are the most common signs of a fake payslip?
Producer metadata that doesn't match the claimed employer's payroll system, font metric inconsistencies between fields, net pay that doesn't reconcile with gross minus deductions, AI-generation noise patterns, and YTD figures that don't align with the pay-period sequence.
Can you detect AI-generated payslips?
Fully synthetic payslips from generative AI tools carry spectral signatures and noise patterns that real payroll-system PDFs don't have. TamperCheck's CV layer catches these even when the payslip passes visual inspection.
How much does payslip fraud detection cost per loan?
Per-document pricing. Most loan applications include 1-3 payslips, so verifying a whole application costs a fraction of any single fraudulent loan that would otherwise be approved.
Does payslip fraud detection slow down loan origination?
Verdicts return fast enough that checks complete while the applicant is still on the next application step. Clean payslips can be auto-approved at the document level, so manual review only sees the suspicious cases.
Can payslip fraud detection replace calling the employer?
It verifies the document, not the employment relationship. Some lenders still confirm employment directly for high-value loans, but checking the payslip first catches most fraud at the document level - before anyone spends time on a call.
Can it work with payslips from any country?
The forensic checks are document-agnostic - PDF structure, font metrics, producer metadata, and AI-generation signatures apply equally everywhere. Country-specific tax-line validation kicks in automatically when the document is classified.
Or skip the call entirely
We'll match your first top-up in credits, so you can run your own documents through it now.