Afghan Identity Verification — Pashto and Dari Names, Documents, and Records

Ariana Nexus provides Afghan identity verification and document authentication — Pashto and Dari names, scripts, and records, across 24 Afghan languages. An Afghan name romanizes a dozen ways; to a system built for one spelling, the real person reads as a mismatch.

Why does one Afghan name become twelve records?

It fails them twice.

Identity systems assume a trusted registry, one spelling per name, and clean biometrics. For Afghan populations none of that holds: names romanize a dozen ways and rarely carry a fixed surname, documents are in Perso-Arabic script, the civil registry is under contested control, and many arrived with no documents at all.

An uncalibrated system fails in both directions — it rejects the real person and clears the fraudulent document, and neither failure announces itself. Ariana Nexus calibrates the system for the population — names, scripts, documents — so neither is decided by a spelling.

Failure one — the real person
The real person
Uncalibrated identity system
Wrongly rejected
A refugee excluded from the benefit, status, or service they are entitled to.
Failure two — the fraudulent document
A fraudulent document
Uncalibrated identity system
Wrongly accepted
Fraud left undetected in exactly the records that matter most.
With Ariana Nexus calibration
The real person — verified
The fraudulent document — caught
eIDAS 2.0
A digital identity wallet in every EU hand by end-2026.
ISO/IEC 30107
The presentation-attack-detection bar, now facing deepfakes.
Transliteration Variance Index
Our name-matching measure.

Even the passport standard concedes the spread: under ICAO Doc 9303, the machine-readable zone and the visual zone of the same passport may carry two different Latin renderings of one name. The Transliteration Variance Index quantifies that spread — and its measured effect on matching.

The Evidence Ledger

What identity failure measurably costs

Three findings from government, standards-body, and peer-reviewed sources. Each is a number a board can act on.

1 in 5
Within a single facility, patient-record match rates can run as low as 80 percent — up to one in five patients unmatched to their chart.
The Pew Charitable Trusts · Enhanced Patient Matching · 2018
42%
Of 3.8 million suspicious-activity reports filed in 2021, the share that was identity-related — $212 billion in reported suspicious activity.
FinCEN Financial Trend Analysis · January 2024
10–100×
Higher one-to-one face-match false-positive rates for Asian and African American faces, across 189 algorithms tested.
NIST FRVT Part 3 · NISTIR 8280 · 2019
Security
ISO/IEC 27001-aligned. Zero security incidents.
Trust Center →
Privacy by design
Data minimization, purpose limitation, no surveillance overreach.
Trust Center →
Fairness
Subgroup-fair false-rejection rates — the real person is not wrongly excluded.
Detection only
Verification and detection only. Never forgery or spoofing tradecraft.

Full assurance documentation in the Trust Center.

What Ariana Nexus authenticates

One population's documents — across every form they take, from a national ID to an academic transcript to a handwritten record a generation old.

Identity documents
Tazkira, e-Tazkira, passports, and national identity documents — read against authentic-document references.
Vital records
Birth, marriage, divorce, and death certificates — names, dates, and seals verified across script and spelling.
Education credentials
Academic transcripts, diplomas, degrees, and professional certifications — verified for credential evaluators, universities, and licensing bodies.
Civil & legal documents
Affidavits, court records, and property and land documents — authenticated with Afghan legal-document expertise.
Benefit & status claims
Eligibility and benefit claims and status applications — the identity behind the claim resolved correctly.
Historical & handwritten records
Older civil records and manuscripts — where script, hand, and era must be read, not just scanned.
Medical records
Medical and health records — verified for benefits, disability, immigration-medical, and continuity-of-care decisions, with Afghan clinical-document and name expertise.
Migration & immigration documents
Visas, permits, residency and immigration case-file documents — authenticated for immigration and border decisions.
Related: Asylum & Resettlement Operations →

If it carries an Afghan name in an Afghan script, Ariana Nexus can read it — and tell you whether it is real.

Media, platform, and content-integrity work → Trust, Safety & Content Integrity
Healthcare

Identity in the health system

A transliterated name is not a data-quality nuisance. It is a patient-safety event waiting to file itself — one new spelling, one new medical record.

Master patient indexes match on the exact fields Afghan names destabilize — spelling, surname structure, field order. A patient registered as Mohammad at the ED, Muhammed at the lab, and Mohamed at the payer is three records: fragmentary histories, repeated tests, mis-routed results, a denial trail. One Texas system reported 2,488 records named Maria Garcia, 231 sharing a birth date — for transliterated names the collision space is wider and the variance is structural. The obligations are in the register below; what is missing is the calibration layer that meets them for this population.

A health system that cannot resolve a transliterated name to one patient cannot meet the obligations it already carries.

Where Programs Stand

The Identity Readiness Ladder

Five levels separate a program that cannot see this population from one whose decisions survive audit. Most programs discover they are lower than they assumed.

L1
UnscoredThe population is invisible to the metrics. No one can say what the false-rejection rate is, because no one has measured it.
L2
Aware◆ Most programs sit hereFailures are known anecdotally — a denied claim, a duplicate file, a flagged document — but nothing is quantified or owned.
L3
MeasuredMatch rates, duplicate rates, and false rejections are broken out by population and script — the gap has a number.
L4
CalibratedName, script, and document handling are tuned to the population; validation gates govern every decision in the pipeline.
L5
Audit-gradeSubgroup parity is documented, evidence is retained, and every identity decision is explainable to a regulator, an auditor, or a court.

The Identity Resolution Review places your program on this ladder — and prices the distance to Level 5.

What Partnership Looks Like

What you receive

The Afghan Identity Resolution Standard™, applied.
Names, documents, and biometrics resolved correctly for Afghan populations.
Afghan name transliteration and matching, calibrated.
The dozen-spelling problem solved, with the Transliteration Variance Index behind it.
Document authentication across every class in the grid above.
Assessed against authentic-document references.
Biometric presentation-attack detection to ISO/IEC 30107.
Spoofs and deepfakes caught, with subgroup-fair false-rejection rates.
eIDAS 2.0 and NIST SP 800-63 assurance-level alignment, documented.
A fairness and false-rejection audit, with privacy and civil-liberties safeguards.
Data minimization, purpose limitation, no surveillance overreach.
Human adjudication for edge cases.
People who read the document and the name, behind every automated decision.

What you receive is not a black-box score. It is the real person verified — and the fraudulent document caught.

Who leads this capability

Wasil Peroz
Practice Leader, Institutional Law
Leads the practice's identity-proofing, document-authentication, and biometric and media-verification work, and the authentication standards that govern it across Afghan names, documents, and media.
Maryam Safi
Principal, Healthcare Research
Leads privacy, fairness, and cultural-compliance governance across the capability — subgroup-fairness review, civil-liberties safeguards, and the protection of legitimate Afghan voices — and holds the CCB Sign-Off Mark.
B.A. | Cornell University
The Cultural Authentication Bench
Human Intelligence Collective · 24 Afghan languages
The bench of Afghan subject-matter experts, native authenticators, and cultural authorities across all 24 Afghan languages who provide the human ground-truth and cultural validation behind every decision. Represented by function, not by name — for the security of the people who do the work.

Questions from compliance, identity, and health-information teams

Can you verify a tazkira, the Afghan national ID?

Yes. The tazkira is the Afghan national identity document, in both its paper booklet and electronic (e-tazkira) forms. Verification covers the document itself, the name as written in Pashto or Dari script, and the transliteration that reached your record. Ariana Nexus does verification and detection only.

How do you match Afghan names across different spellings?

By resolving the script rather than the spelling. One name written محمد can romanize a dozen ways — Mohammad, Muhammad, Mohammed, Mohamed and more — and each spelling can open its own record. Matching works back from the original Pashto or Dari orthography to one verified identity instead of guessing between English variants.

What if the issuing office in Afghanistan no longer functions?

That is a normal condition of Afghan casework, not an exception. Where an issuing authority cannot be queried, the assessment shifts to what can be evidenced — document features, internal consistency, and corroborating records — and is reported with a stated confidence level. Ariana Nexus reports what it can and cannot establish rather than asserting a verification the record will not support.

Do you work in both Pashto and Dari script?

Yes. Pashto and Dari are both written in Arabic script, but they differ in letters, orthography, and naming convention, and Pashto carries additional characters Dari does not. Reading a name correctly means knowing which language produced it — across 24 Afghan languages, not only the two.

How do you tell a genuine Afghan document from a fraudulent one?

Through calibration against genuine documents from the same population, period, and issuing authority. An uncalibrated system produces both failures at once — it wrongly rejects real people and wrongly accepts fraudulent documents. Findings carry a confidence level rather than a binary verdict the evidence will not support.

Do you provide forgery or spoofing services?

No. Verification and detection only — never forgery or spoofing tradecraft. That boundary is fixed, and the full assurance documentation sits in the Trust Center.

Request an Identity Resolution Review

Briefings are conducted under NDA, in Washington, D.C. or virtually.

Request a confidential briefing