Document verification

Document verification confirms that a government-issued identity document is authentic, readable, and acceptable for your flow, and extracts the fields you need for onboarding and compliance.

What it does

1. Guides the user to capture or upload the document
2. Reads the contactless chip where the document carries one; see Chip reading
3. Runs authenticity and quality checks (security features, tampering, copy / screen attacks, score thresholds); see Document capture checks
4. Applies your country and document-type filters
5. Enforces minimum age when configured
6. Produces a step result (PASSED, FAILED, or review statuses) with optional structured reasons

Chip reading

Many modern identity documents carry a contactless chip: most current passports, and biometric national identity cards and driving licences in a growing number of countries. Where the presented document has one and the capture device can reach it, Zyphe reads the chip in addition to the printed document.

The chip matters because its contents are signed by the issuing authority. Fields read from it are not a transcription of what a camera saw, so a forged or altered printed surface does not survive comparison with them, and the portrait can be taken from the chip rather than from the photograph on the card.

Chip reading supplements the image pipeline, it does not replace it. A document is still checked for tampering, screen and paper capture, template match and MRZ consistency, because not every document carries a chip and not every capture context can reach one. Where no chip is present or it cannot be read, verification proceeds on the printed document and the checks below.

Document capture checks

When a document is captured or uploaded, the platform runs automated authenticity, quality, and data-consistency checks before accepting it. Failures surface as structured codes on the step result (data.dv.reasons[]); see Statuses and codes.

Check	Description
Tamper check	Detects if the ID's main portrait was physically tampered.
ID alteration check	Validates whether there are any foreign objects on the document.
Alignment	Flags when the picture is badly aligned and the alignment/cropping correction could not produce a usable crop, and when the cropped capture is too close to the image edges.
Screen ID liveness	Confirms whether the document was captured from a screen rather than the physical plastic document.
Paper ID liveness	Confirms whether the document was captured from a printed paper rather than the physical plastic document.
Voided ID check	Flags IDs that have been intentionally voided, such as those with cut corners, punched holes, or visible void marks. These alterations indicate the ID is no longer valid for use.
Fake ID check	Searches for discrepancies in the document layout and other signs of fraud.
Template check	Checks whether the document layout matches a supported document template (text positions and presence of security features).
MRZ cross-check	Checks whether data extracted from the MRZ matches data extracted by OCR.
Expiration date validity	Confirms that the expiration date is valid.
Document type side cross-check	Checks front/back consistency of the document (for example, if both images are the front of the ID).
Issue date validity	Confirms that the document issue date is valid.
Barcode presence and format	When a barcode is expected on the document template, checks that it is present and matches the expected format.
Image quality check	Flags glares above a threshold and low-resolution images.

These checks contribute to the document authenticity score and step outcome. You can tune the minimum acceptable document score and related filters in the flow builder.

Where it sits in a flow

Document verification is usually the first identity step. Pair it with liveness so you know a live person matches the document portrait. Other steps (PoA, phone, forms) can follow depending on risk appetite.

Document verification → Liveness → (optional PoA / form / phone) → Result + webhook

Outcomes you should handle

Outcome	What to do
PASSED	Continue the flow or accept the identity evidence
FAILED + reasons[]	Map codes to UX (retry, different document, support), statuses and codes
Manual review	Keep the user pending until review completes, manual review

Configure it

Operational settings (thresholds, filters, labels) live in the flow builder:

→ Document verification guide · Create a flow

Related

• Antifraud controls, where document checks sit next to liveness, KYB, AML, and TM
• Verification process, public link vs SDK
• Webhook payload: DV
• Reusable credentials, reuse a prior verification when eligible