Extract data from Form I-9 documents

Every US employer has to verify that a new hire can legally work, and the USCIS Form I-9 is the record that proves it was done. HR teams complete one for each employee and keep it on file for audit, and a company onboarding 300 people a year accumulates a drawer of them that an Immigration and Customs Enforcement audit can demand on three days notice. The form has two halves: Section 1, which the employee completes with their name, date of birth, and attestation of citizenship or work-authorization status, and Section 2, which the employer completes by recording the identity and work-authorization documents the employee presented, drawn from List A or a combination of List B and List C. The hard part is the document logic and the timing rules. Section 2 accepts either one List A document (such as a US passport, which proves both identity and work authorization) or one List B identity document plus one List C work-authorization document, and a valid I-9 cannot mix them incorrectly. Each presented document is recorded with its title, issuing authority, number, and expiration date, and a work authorization that expires triggers reverification. The form carries strict deadlines: Section 1 by the first day of work, Section 2 within three business days, and a missed date is the most common audit finding. The version of the form matters, since USCIS retires editions. Talonic reads the Form I-9 and returns the Section 1 employee attestation, the citizenship or work-authorization status, and the Section 2 documents with their list category, numbers, and expiration dates. An HR team audits employment eligibility and flags reverification dates from structured data instead of reading each form. An I-9 completed 2026-06-09, with Section 2 finished 2026-06-11 by a US employer, records a Permanent Resident Card issued by USCIS expiring 2031-03-14 for the audit file.

What gets extracted from Form I-9 documents

How extraction works for Form I-9 documents

Form I-9 records arrive as portal PDFs, signed printouts, and scans, and HR keeps them for audit rather than downstream processing. Talonic reads the form and maps it to the employment-eligibility schema in the Field Registry, which separates the Section 1 employee attestation from the Section 2 employer document record. The citizenship or work-authorization status is captured from Section 1, and the Section 2 documents are captured with their list category (List A, or List B plus List C), title, issuing authority, number, and expiration date. The List A versus List B and C logic is checked so an invalid combination is flagged. Expiration dates are surfaced so an expiring work authorization triggers reverification, and the Section 1 and Section 2 completion dates are checked against the first-day and three-business-day deadlines. Every value returns with a confidence score and pixel-region provenance under DIN SPEC 91491 conformity, so an HR auditor can verify a document entry against the source form.

Sample extraction

A completed USCIS Form I-9 with a List A document

{
  "employee_name": "Lukas Bauer",
  "date_of_birth": "1990-03-14",
  "citizenship_status": "lawful_permanent_resident",
  "section2_documents": [
    {
      "list": "A",
      "title": "Permanent Resident Card",
      "issuing_authority": "USCIS",
      "number": "MSC2090000000",
      "expiration_date": "2031-03-14"
    }
  ],
  "section1_completed": "2026-06-09",
  "section2_completed": "2026-06-11",
  "reverification_required": false
}

Frequently asked