Extend has launched Extend Verify, a document fraud detection API aimed at developers and AI agents that rely on documents to make decisions. The company notes that AI has made forgeries imperceptible to human reviewers easy to produce, creating risk for lending, payments, insurance, and background checks. Verify is designed to let agents flag suspicious inputs before acting on them.
The API examines visual content, file structure, and revision history, combining specialized models, forensic checks, and agentic analysis. It looks for signs of AI-generated documents, localized tampering down to single-character edits, conventional digital editing, and changes made after digital signing. Verify returns risk scores, explanations, and traceable evidence so businesses can decide which documents need further review.
In a benchmark evaluation, Verify detected 96.3% of fraudulent documents at a 4.0% false-positive rate. Example use cases include flagging altered income documents before credit decisions, surfacing modified invoices before payment, and identifying fabricated receipts before claim settlement. Developers can use Verify via API alone or add it to existing Extend pipelines that parse and extract document data. This report is based on a single vendor announcement; no independent comparison or conflicting source was available.