Digital Identity Verification in Cross-Border Payments

1. Executive Summary

Identity verification sits at the heart of every cross-border payment transaction. Regulators require money transfer operators (MTOs) to verify who their customers are before processing transfers, primarily to prevent money laundering, terrorist financing, and fraud. For the customer, the verification experience is often the first interaction they have with a provider, and it sets the tone for the entire relationship.

Traditional identity verification methods, relying on in-person document inspection at a physical branch, are being rapidly replaced by digital solutions. Document scanning, optical character recognition (OCR), biometric matching, and electronic verification databases now enable customers to verify their identity from a smartphone in minutes rather than days. These advances have been transformative for digital-first remittance services, reducing onboarding friction and expanding access.

Yet digital identity verification also presents challenges, particularly for customers from developing countries whose identity documents may differ from the standards assumed by verification technology, and for providers seeking to balance regulatory requirements with the imperative to serve underbanked populations. This white paper examines the landscape of digital identity verification in cross-border payments, with specific attention to the challenges and opportunities relevant to the UK-to-Gambia corridor.

2. Why Identity Matters in Cross-Border Payments

Identity verification in cross-border payments serves multiple interrelated purposes:

2.1 Regulatory Compliance

The UK's Money Laundering Regulations 2017 (MLRs) require that regulated firms verify the identity of their customers before establishing a business relationship or carrying out occasional transactions above applicable thresholds. For MTOs, this typically means verifying every customer before their first transfer. The purpose is to ensure that the financial system cannot be used to launder the proceeds of crime or finance terrorism.

2.2 Fraud Prevention

Identity verification is the first line of defence against fraud, including identity theft (where a criminal uses someone else's identity to send or receive money), account takeover, and application fraud. Robust verification reduces financial losses for providers and protects innocent consumers whose identities might otherwise be misused.

2.3 Sanctions Compliance

Accurate identity data is essential for effective sanctions screening. Without reliable customer identification, it is impossible to screen accurately against sanctions lists, which could result in either processing transactions for sanctioned individuals or generating excessive false positive alerts that degrade the customer experience.

2.4 Customer Trust

Paradoxically, customers who might initially find verification inconvenient ultimately benefit from and value it. A provider that takes verification seriously signals that it is legitimate, regulated, and invested in protecting its customers, factors that are particularly important in diaspora communities where trust is paramount.

3. Traditional Verification Challenges

Before the advent of digital identity solutions, MTOs relied on in-person document inspection. A customer would visit a branch with their passport or driving licence, and a staff member would manually check the document's authenticity and record the customer's details. This approach presented several challenges:

4. Document Verification Technology

Modern document verification uses a combination of hardware (smartphone cameras) and software (AI-powered analysis) to authenticate identity documents remotely.

4.1 The Verification Process

A typical digital document verification flow involves the following steps:

1. Document capture: The customer photographs their identity document using their smartphone camera. Guidance overlays help ensure the image is properly framed, focused, and lit.
2. Document classification: AI models identify the document type (passport, driving licence, national ID card) and the issuing country.
3. OCR extraction: Optical character recognition extracts text data from the document, including name, date of birth, document number, expiry date, and machine-readable zone (MRZ) data.
4. Authenticity checks: The system analyses visual security features (holograms, microprinting, UV patterns), checks for signs of digital manipulation, and validates MRZ checksums against extracted data.
5. Database verification: Extracted data is cross-referenced against government databases (where available) and fraud databases to confirm validity.

4.2 Supported Document Types

Leading verification providers support thousands of document types from over 200 countries. For the Gambian diaspora in the UK, the most commonly presented documents include:

• UK passport: The gold standard for verification, with excellent machine readability and extensive security features. Many naturalised Gambian-British citizens hold UK passports.
• UK driving licence: Widely accepted and well supported by verification systems.
• Gambian passport: The most common document for Gambian nationals. Newer biometric passports are well supported, but older hand-written passports can present challenges for automated systems.
• Gambian National Identity Card: Increasingly used but less well supported by international verification databases than passports.
• UK Biometric Residence Permit (BRP): Held by Gambian nationals with settlement or limited leave to remain. Well supported by verification systems.

4.3 Accuracy and Error Rates

5. Biometric Solutions

Biometric verification adds a second layer of assurance by confirming that the person presenting the document is the person pictured on it.

5.1 Facial Recognition (Selfie Match)

The most widely deployed biometric in remote identity verification is facial recognition, or "selfie match." The customer takes a selfie photograph, and AI algorithms compare the facial features to the photograph on the identity document. Modern facial recognition systems achieve match accuracy of over 99% under good conditions.

5.2 Liveness Detection

To prevent spoofing (where a fraudster presents a photograph or video of the genuine customer rather than their own face), verification systems incorporate liveness detection. This may involve:

• Active liveness: The user is asked to perform an action, such as turning their head, blinking, or smiling, that a photograph cannot replicate.
• Passive liveness: AI analyses subtle visual cues (texture, depth, reflection patterns) to distinguish a live face from a photograph or screen reproduction, without requiring user action.

5.3 Bias and Fairness

Facial recognition technology has been subject to well-documented concerns about bias, particularly lower accuracy rates for people with darker skin tones. For MTOs serving the African diaspora, this is not an abstract concern but a practical challenge that can lead to legitimate customers being incorrectly rejected.

Research by the National Institute of Standards and Technology (NIST) has shown that many facial recognition algorithms exhibit demographic differentials, with higher false rejection rates for African and African-descended populations. However, the gap has narrowed significantly as training datasets have become more diverse and algorithms more sophisticated. Leading verification providers now publish demographic performance metrics and actively work to minimise bias.

"If your identity verification system works flawlessly for European passport holders but fails for Gambian passport holders, you have not built a global verification system. You have built a system that excludes the people who need it most."

6. eKYC Systems and Digital Identity

Electronic Know Your Customer (eKYC) goes beyond document and biometric verification to encompass a broader ecosystem of digital identity sources.

6.1 Data Sources

UK-based MTOs can draw on a rich ecosystem of electronic data sources to verify customer identity:

• Credit reference agencies: Experian, Equifax, and TransUnion hold address history and identity data for most UK residents.
• Electoral roll: Public register data provides address verification for registered voters.
• Government databases: DVLA (driving licence), HMPO (passport), and DWP (national insurance) databases can be accessed through authorised intermediaries.
• Telecommunications data: Mobile phone contract data can verify name and address.
• Bank account verification: Open banking APIs can confirm account ownership.

6.2 Multi-Source Verification

Best practice in eKYC combines multiple data sources to build a composite identity confidence score. A customer whose name and address match across credit reference data, electoral roll, and a document check achieves a high confidence level. This multi-source approach is more robust than relying on any single verification method.

6.3 Challenges for Diaspora Customers

Diaspora customers, including the Gambian community in the UK, can face specific challenges with electronic verification:

• Thin credit files: Recent arrivals to the UK may have limited or no credit history, reducing the effectiveness of credit reference checks.
• Shared housing: Multiple family members at the same address can complicate address verification.
• Name variations: Different spellings or orderings of names across documents (e.g., compound names, patronymics) can cause mismatches in automated systems.
• Limited digital footprint: Newer residents may lack the digital traces (utility bills in their name, registered voter status) that electronic verification systems rely on.

7. Gambian Identity Systems

Understanding the identity document landscape in The Gambia is relevant for MTOs that need to verify recipient identities or onboard customers who may present Gambian documents.

7.1 National Identity Card

The Gambia Identity Management System (GIMS), managed by the Department of Immigration, issues national identity cards to Gambian citizens and residents. The newer biometric cards, introduced in phases from 2019, include chip-based storage of biometric data (fingerprints and facial image), security features, and a unique identification number. However, coverage is not yet universal, with an estimated 60-70% of the adult population holding a current national ID card.

7.2 Gambian Passport

The Gambian e-Passport (biometric passport), introduced in 2018, meets ICAO (International Civil Aviation Organization) standards with a polycarbonate data page, MRZ, and embedded chip. The biometric passport is well supported by international verification systems. However, many Gambians still hold older, non-biometric passports, which are less reliably verifiable by automated systems.

7.3 Birth Registration

Birth registration remains a significant challenge in The Gambia, with UNICEF estimating that only 53% of children under five had their births registered as of 2023. The absence of a birth certificate can create difficulties for individuals seeking to obtain other identity documents, creating a chain of identity exclusion that extends into adulthood.

8. Balancing Security with Inclusion

The central tension in identity verification for cross-border payments is between security (ensuring only legitimate customers are onboarded) and inclusion (ensuring that verification requirements do not exclude legitimate customers who happen to have non-standard documentation or limited digital footprints).

8.1 The Exclusion Risk

Overly rigid verification requirements can exclude exactly the populations that digital remittance services are designed to serve. If a Gambian national in the UK cannot pass automated verification because their passport is an older format, their credit file is thin, and their name is spelled differently across documents, they are pushed towards cash-based, in-person services, which are typically more expensive and less convenient.

8.2 Risk-Based Verification

The solution lies in risk-based verification approaches that apply appropriate levels of scrutiny based on the risk profile of the customer and the transaction:

8.3 Inclusive Design Principles

MTOs can improve inclusion without compromising security by adopting the following principles:

• Support a wide range of documents: Accept not just passports and driving licences but also BRPs, national ID cards, and travel documents.
• Invest in diverse training data: Ensure that document recognition and facial matching algorithms are trained on diverse datasets that include African documents and faces.
• Provide human fallback: When automated verification fails, offer a manual review pathway rather than simply rejecting the customer.
• Offer multilingual guidance: Provide clear instructions in the languages spoken by target communities (English, Mandinka, Wolof).
• Test with real users: Conduct usability testing with actual diaspora customers to identify and fix friction points.

9. Future Directions

Several emerging technologies and regulatory developments are likely to reshape identity verification in cross-border payments over the coming years:

9.1 Digital Identity Wallets

The UK government is exploring digital identity frameworks that would allow citizens to carry verified identity credentials on their smartphones, potentially enabling instant, high-confidence verification without the need to photograph physical documents.

9.2 Reusable Identity

The concept of "verify once, use many" is gaining traction. If a customer completes identity verification with one regulated provider, that verification could, in principle, be shared with other providers (with the customer's consent), reducing duplication and friction. Open banking and digital identity standards are laying the groundwork for this approach.

9.3 AI-Powered Continuous Verification

Rather than verifying identity only at onboarding, AI systems are increasingly performing continuous identity assurance, analysing behavioural patterns (typing speed, device usage, transaction patterns) to confirm that the person using the account is the verified customer. Anomalies trigger step-up verification rather than blocking the transaction outright.

9.4 Decentralised Identity

Blockchain-based decentralised identity systems, where individuals control their own identity data rather than relying on centralised databases, offer a long-term vision for identity management that could be particularly powerful for populations with limited access to traditional identity infrastructure.

10. Conclusion

Digital identity verification has made cross-border payments faster, more secure, and more accessible. For the Gambian diaspora in the UK, the ability to verify their identity through a smartphone app in minutes, rather than visiting a branch with physical documents, has opened up access to cheaper, faster, and more convenient money transfer services.

But technology alone is not sufficient. The verification systems that serve this community must be designed with an understanding of the specific challenges that diaspora customers face: non-standard documents, name variations, thin credit files, and the biases that can be embedded in facial recognition algorithms. Inclusive design, human fallback processes, and a risk-based approach that avoids unnecessary exclusion are all essential.

FRS Money is committed to building verification processes that are both robust and inclusive. We invest in verification technology that works for our community, because we believe that security and inclusion are not opposing forces but complementary goals.

Sources: NIST Face Recognition Vendor Test (FRVT); FCA Guidance on Electronic Identity Verification; JMLSG Guidance Part I; ICAO Doc 9303 (Machine Readable Travel Documents); UNICEF State of the World's Children 2024; Gambia Immigration Department; UK Digital Identity and Attributes Trust Framework; FRS Money internal verification data.