One fraudulent onboarding can undo months of compliance work.
Identity fraud is increasingly sophisticated. AI-generated IDs, photo-swapped passports, and digitally composited documents now pass visual inspection with ease. By the time a fraudulent customer is flagged - often during an audit or incident - the damage is already done. TamperCheck gives your onboarding team a forensic layer that catches what the human eye misses.
$48B+
lost globally to identity fraud each year
1 in 5
fraudulent documents passes initial manual review
~1 minute
to get a verdict on any identity document
The problem
What fraud looks like in kyc & identity verification
AI-generated identity documents
Generative AI tools now produce convincing passports and ID cards that include realistic fonts, holograms, and MRZ data. They pass visual inspection but fail on pixel-level forensics - noise patterns, compression artifacts, and spectral analysis expose them.
Photo zone substitution
Legitimate documents with a replaced portrait zone. The original document is genuine; only the photo has been swapped. Boundary contrast analysis, sharpness gradients, and compression mismatch in the portrait region reveal the edit.
Date and field tampering
Date of birth, expiry dates, and ID numbers are the most commonly altered fields. Font inconsistencies, spacing anomalies, and metadata mismatches between the visual layer and embedded data expose these changes.
How TamperCheck helps
A forensic verdict in about a minute
TamperCheck runs 130+ forensic checks across visual, structural, and AI-analysis layers on every upload - whether it's a high-resolution scan, a phone photo, or a PDF. The result is a plain-English verdict with pinpointed findings, so your compliance team knows exactly what to look at, not just that something is flagged.
What we check
- MRZ line integrity and character consistency
- Portrait zone boundary contrast and sharpness
- Hologram and security overlay plausibility
- Font metrics and spacing across all fields
- AI-generation spectral and noise pattern analysis
- Metadata consistency between document layers
Documents we verify
Common document types in kyc & identity verification
Frequently asked
Common questions about kyc & identity verification
How does AI KYC detect deepfake identity documents?
AI KYC document forensics uses spectral analysis, noise pattern detection, and metadata inspection to identify documents generated by generative AI tools. Real scanners and printers produce characteristic artifacts that AI-generated documents lack - TamperCheck's CV layer is trained on these signatures and catches synthetic IDs that pass every visual check.
Can TamperCheck replace my existing KYC provider?
No - it's designed to slot in alongside providers like Onfido, Jumio, Persona, and Sumsub. Keep your provider for face matching, watchlist screening, and OCR; add TamperCheck as the document forensics layer that catches forged and AI-generated IDs.
Does it work on identity documents from any country?
Yes. TamperCheck analyses the forensic properties of the document - PDF structure, font metrics, pixel-level signals, AI-generation signatures - not country-specific templates. This means it works on passports and IDs from any issuing authority, including documents it has never seen before.
What's the difference between liveness detection and document forensics?
Liveness confirms the person submitting a selfie is physically present (not a photo, video, or deepfake). Document forensics confirms the submitted document is authentic. They're parallel checks on different attack surfaces - a complete KYC stack uses both.
How does pricing work for KYC document verification?
$0.50 per document, no subscription, no minimums. Most KYC workflows submit 1-3 documents per applicant, so per-applicant verification cost is $0.50-$1.50. There's a free tier for evaluation.
See it working on your documents
Start with $5 in free credits - no contract, no card required. Upload your first document and get a verdict in about a minute.