AI-generated fakes
Our own models find the fingerprint image generators leave in grain, print and light.
THE HARBOUR KITCHEN
Looks genuine to a personTasting menu x6510.00
Chablis x3186.00
Service90.00
TOTAL€786.00
NewHolistic claim review is live →
Document fraud detection for every payout.
Docklands AI is document fraud detection anchored on the financials behind every payout. It catches the AI fakes, traditional fraud, fraud networks and claims inflation that metadata checks miss.
app.docklands.ai/claims/CLM-1407
INV-20931.jpgquote_northgate.pdfphoto_07.heiccard_receipt.png+19
NORTHGATE AUTO REPAIRS
Unit 4, Harbour Rd · VAT IE1234567T
14/09/2026INV-20931
Brake pads (front)289.00
Brake discs x2389.00
Labour 3.5h420.00
Diagnostic95.00
Subtotal1,193.00
VAT 23%274.39
TOTAL EUR1,467.39
Total re-written
Cloned digits
No business here
+46% vs market
22 supporting files checked+14
Checks 0 / 214Fraud risk
INV-20931 · 23 files on claim
Analysing…Request · illustrative# send every file on a claim curl -X POST https://api.docklands.ai/v1/documents \ -H "Authorization: Bearer $DOCKLANDS_KEY" \ -F "files=@INV-20931.jpg" \ -F "files=@quote_northgate.pdf" \ -F "files=@photo_07.heic" # or forward to your claims inbox # claims@acme.docklands.ai
Response · illustrative{ "claim": "CLM-1407", "risk": { "score": 92, "level": "high" }, "findings": [ { "type": "ai_generated", "confidence": 0.97 }, { "type": "tampering", "field": "total" }, { "type": "cloned_region", "count": 2 }, { "type": "address_no_business" }, { "type": "inflation", "line": "labour", "above_market": 0.46 }, { "type": "network_link", "claims": 5 } ], "checks_run": 214, "false_positives_cleared": 1 }
88%AI-generated receipts caught in a US production benchmark
200+checks on every invoice
43smedian time to a verdict
<30 minto integrate by API or email
Capabilities
Going beyond just metadata.
We build a foundation on the payment, this is what is unlocked →
Manual tampering
Correction fluid, handwritten digits, an overwritten total. Drag to compare the photo with what our model sees.
NORTHGATE AUTO
Subtotal1,193.00
VAT274.39
TOTAL1,467.39
paid 14/9 ✓NORTHGATE AUTO
SubmittedModel viewSubtotal1,193.00
VAT274.39
TOTAL1,467.39
paid 14/9 ✓Correction fluidHandwritten "4"Digital edits
Cloned digits and pasted lines, found pixel by pixel.
289.00389.00
Identical pixels: "89.00"
200+ checks per invoice
Every number, date, bank detail and address.
- VAT number formatpass
- Line items add uppass
- Date is a trading daypass
- IBAN checksumpass
- Phone number in servicepass
- Business at addressfail
- Invoice number sequencepass
- VAT rate matches categorypass
- VAT number formatpass
- Line items add uppass
- Date is a trading daypass
- IBAN checksumpass
- Phone number in servicepass
- Business at addressfail
- Invoice number sequencepass
- VAT rate matches categorypass
Inflation
Every line priced against the market.
Brake pads+76%
Discs x2+34%
Labour+46%
Diagnostic+5%
Holistic claim review
Every file cross-referenced. Catches what no single document shows, and clears false positives.
Invoice€1,467.39
5 findings
5 findings
Quote€980.00
Invoice +50%
Invoice +50%
Photos · 1116/09
After repair
After repair
Card€1,467.39
Matches
Matches
2 cross-reference issues1 false positive cleared
Fraud networks
One receipt unlocks every claim that shares its vendor, bank account, phone or device.
What links them
Details the receipt shares with other claims.
- Shared IBAN ••44713 claims
- Shared phone ••2182 claims
- Same vendor2 claims
- Same capture device1 claim
- Linked exposure€435.6K
Works with what you have
Send documents by API, email, or straight from SharePoint and Google Drive. Results go back to your claims system, case manager or inbox.
REST APIEmail intakeSharePointGoogle DriveWebhooksMFASSOCSV export
See what Docklands AI finds in your own documents.
Your first 1,000 documents are checked free.
How it works
From one document to a decision. In seconds, not days.
Every file on a claim goes through the same four stages. Your team only sees what needs a human.
Ingest
API, email forwarding or upload. Receipts, invoices, PDFs, phone photos: one file or a hundred.
Inspect
Custom models for AI fakes, tampering and edits, plus 200+ checks on every invoice.
Cross-reference
A holistic review compares every file on the claim and clears flags the full picture explains.
Decide
A risk score with evidence pinned to each document, sent to your claims system or inbox.
ROI calculator
How much could you save?
Calculate your potential savings from document fraud detection with our interactive ROI calculator.
Reports & integrations
Alerts to your inbox.
Send documents by API, email or straight from SharePoint and Google Drive. Build the report you need with drag and drop, and get it in your inbox the moment fraud is found.
Easy ingestion
Connect the way documents already reach you. No new upload portal for your team.
REST APIEmailSharePointGoogle Drive
Drag-and-drop reports
Choose what goes in: risk score, findings, evidence images, network maps and inflation. No code and no IT ticket.
Alerts in your inbox
When fraud is detected, your report arrives by email with the evidence attached, ready for the claim file.
Security
Built for claims data.
Your documents stay protected, auditable and where you need them.
Multi-factor login
MFA with backup codes, trusted devices and guided recovery.
US or EU data residency
Choose where your documents are processed and stored.
Full encryption
Every document and every finding is encrypted.
Custom data retention
Keep documents as long as your policy requires, and no longer.
SOC 2 Type I
Independently audited security controls.
ISO 27001
Certified information security management.
Industries
Anywhere money moves on a document.
Insurance claims
Repair invoices, medical bills and receipts behind motor, property, travel and health claims.
Explore insurance claims →- AI-generated repair invoices
- Inflated garage and contractor bills
- Quote vs invoice mismatches across a claim
- Vendor rings shared across claimants
FAQ
Questions.
What is document fraud detection in insurance claims?
Document fraud detection analyzes receipts, invoices and proof of payment submitted with insurance claims to identify fake, altered or inflated documents. It catches AI-generated receipts, manually tampered invoices, digitally edited bank statements and inflated repair bills before payouts are made.
How do insurers detect fake proof of payment documents?
Advanced systems use custom models trained on document images to spot AI generation patterns, pixel-level digital edits and physical tampering that metadata analysis misses. They also cross-reference documents across claims to identify shared vendors, bank accounts and devices that signal fraud rings.
Which claim lines does Docklands work for?
Docklands works for any claim line that takes proof of payment: auto, home, renters, warranty and pet insurance. It analyzes repair invoices, medical bills, contractor estimates, vet bills and receipts across property, casualty, health and specialty lines.
How does Docklands fit into an existing claims workflow?
Docklands adds a fraud detection layer without replacing existing claims systems. Documents arrive by API, email forwarding or SharePoint, are checked within seconds, and findings return to your claims system or inbox with evidence pinned to each file.
How is Docklands different from metadata checkers?
Metadata is easy to strip or forge. Docklands runs its own models on document images to detect AI generation, tampering and edits, then runs 200+ checks on invoices and cross-references every file on a claim.
How long does integration take?
Under 30 minutes via REST API, or send documents by email or from SharePoint and Google Drive with no integration.
Does it replace our claims system?
No. Docklands adds a detection layer to the systems you already use and returns findings with the evidence located on each document.
How does it keep false positives down?
After every file on a claim is processed, a holistic review cross-references them and clears flags that the full picture explains.
Still have a question? Ask us on a call. Book a call
Send us the document you don't trust.
We'll show you what's been faked, what's been changed, what's been inflated and who else it's connected to.
Your first 1,000 documents. Checked free.
From the blog
Latest on document fraud.
28 Oct 2026Manual Edit Detection: How to Find Hidden Cuts and Altered Content in Video, Photos and Documents
Software can’t catch every hidden cut or doctored receipt. Learn how to manually detect missing video edits, altered photos, and fraudulent PDFs using visual clues and file-level data before they slip through the cracks.
25 Oct 2026Photoshop Detection: How to Tell If a Photo Has Been Edited
You don't need a forensic lab to spot a doctored photo. Learn how to uncover hidden edits, manipulated lighting, and fake documents using the exact visual checks and free forensic tools professionals rely on to verify images in minutes.
27 Aug 2026Identity Fraud Detection Using Machine Learning, Explained
Discover how machine learning is transforming identity fraud detection. This guide explains how modern algorithms catch what traditional rule-based systems miss, from synthetic identities to manipulated documents.