Home/Guides/Document Fraud Detection

Pillar

Document fraud detection that catches what humans miss.

Document fraud detection is the process of analysing a file - PDF, image, or scan - for evidence of editing, forgery, or AI generation. TamperCheck runs 130+ forensic checks per document across structural, visual, and AI-adjudication layers, returning a plain-English AUTHENTIC or TAMPERED verdict in about a minute.

The problem

Why document fraud detection has to be automated now

Document fraud has changed shape in the last 24 months. The crude Photoshop edits of the 2010s are still around, but they're no longer the threat that matters. AI-generated payslips, bank statements, invoices, and identity documents now pass visual inspection trivially. A bad actor with a $20 subscription can produce a convincing fake in seconds.

Manual review caught roughly 1 in 5 sophisticated edits even before AI generation became commonplace. Today that number is closer to 1 in 10. The math is straightforward: the fakes are getting better faster than reviewers can be trained to spot them.

Automated document fraud detection closes the gap by analysing things humans can't see - PDF font tables, pixel-level noise patterns, compression artifacts, running balance arithmetic, metadata producer fingerprints, and AI-generation spectral signatures. It's not a replacement for a compliance team; it's a forensic layer underneath them.

What we detect

Forensic signals that expose fraud

Structural PDF analysis

TamperCheck inspects the internal structure of every PDF - object streams, font tables, embedded metadata, cross-reference tables - for the fingerprints that editing tools leave behind.

Pixel-level forensics

Error Level Analysis, noise residue, JPEG ghost detection, and compression artifact maps reveal regions of an image or scan that have been edited.

AI-generation detection

Synthetic documents from generative AI exhibit spectral signatures and noise patterns that real scanners and printers don't produce. TamperCheck's CV layer is trained on these signatures.

Arithmetic & logical consistency

Running balances, subtotals, GST calculations, date sequences, and cross-field relationships are all verified. A single edited number breaks the chain.

Producer metadata fingerprinting

Genuine documents have consistent producer metadata (which software created them). Tampered files often show traces of editing tools that the original issuer would never use.

Visual layer vs text layer divergence

When a PDF is edited, the rendered pixels and the underlying text layer can drift apart. TamperCheck detects these mismatches at the character level.

The solution

How TamperCheck delivers document fraud detection at scale

TamperCheck is an API-first document fraud detection tool built for B2B teams that process documents at volume - lenders, insurers, KYC providers, HR teams, and claims adjusters. One endpoint, one upload, one verdict per minute.

Digital PDFsPhone photosScanned hard copiesPhysical stampsHandwritten signatures

How it works

  • API-first - integrate in under 30 minutes; webhooks for async results
  • 130+ forensic checks per document, automatically
  • Plain-English findings that compliance teams can act on
  • Risk score (0–100) calibrated per document type
  • $0.50 per document - no subscriptions, no minimums
  • Zero document storage - files processed ephemerally
  • Works on any document type - bank statements, payslips, IDs, invoices, certificates

FAQ

Common questions

Related

Keep reading

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.