Verify API
Catch tampered and AI-generated documents
Fraud checks inspect one document for AI generation, edits, and post-signature tampering, then return a risk level with evidence located on the page.
What it does
Risk level and summary
Each run returns LOW, MEDIUM, or HIGH risk and a plain-language summary of the findings.
Located evidence
Every signal points to its page, bounding box, PDF objects, or byte ranges so a reviewer can check it.
Two engines
Agentic parses the file and also reads it for arithmetic, sequence, date, and identity conflicts; Standard skips parsing for faster screening.
Under the hood
Detect document tampering and AI-generated fraud before your agents act on it.
AI generated document detection
Specialized models detect and flag fully AI-generated synthetic documents.
AI tampering detection
Detect signs of AI edits within genuine documents, down to changes as granular as a single character.
In-painting forgery
Forensic checks flag regions where original text or images have been erased, replaced, or covered.
Post-signature changes
Digital signature and revision history checks identify edits overlaid after signing.
Benchmarks
Internal fraud benchmark
96.3%Fraudulent documents flagged
Across four fraud types on a challenging internal benchmark.
Internal fraud benchmark
4%False-positive rate at that detection rate
A LOW result does not certify a document as genuine; review flagged evidence before acting.
How to get started
Create an account
Sign up at dashboard.extend.ai. New accounts get 10,000 free credits.
Copy an API key from the Developers page and export it as EXTEND_API_KEY.
Verify is enabled per organization, so request access for your workspace.
Choose how to call Verify
Studio: try Verify in the browser with no code.
SDKs: pip install extend-ai or npm install extend-ai; Java and Go SDKs are also available.
Process your first document
Send a PDF, JPEG, PNG, or WebP file by URL or file ID.
curl -X POST https://api.extend.ai/verify \ -H "Authorization: Bearer $EXTEND_API_KEY" \ -H "x-extend-api-version: 2026-02-09" \ -H "Content-Type: application/json" \ -d '{ "file": { "url": "https://extend-public-files.s3.us-east-2.amazonaws.com/bank_statement_tampered.pdf" } }'Move to production
Send production traffic to POST /verify_runs; the synchronous POST /verify times out after 5 minutes.
Related reading
Use cases
- Financial services
Banking & lending
Check bank statements and income documents before credit decisions.
- Verify launch post
Payments & fintech onboarding
Flag high-risk documents so onboarding scales without more manual review.
- Real estate
Real estate
Catch altered pay stubs, bank statements, and tax returns before a lease or mortgage.
- Verify launch post
Ecommerce
Flag altered seller documents and fabricated receipts in refund claims.
- Verify launch post
HR & staffing
Flag forged diplomas, licenses, and employment records before a background check.
- Verify launch post
Insurance claims
Check the documents behind a claim before it triggers a payout.
