eKYC False Positives: Why Genuine Customers Get Rejected and How Businesses Can Prevent It

Introduction: When Identity Verification Gets It Wrong

Imagine applying for a digital bank account. You upload your MyKad, take a selfie, and complete the identity verification process. Everything appears correct. Your identification document is genuine, your photograph matches your face, and you are the rightful owner of the identity.

However, the system flags your application as suspicious and rejects it.

You try again, but the result remains unchanged.

This is one of the challenges businesses face when implementing Electronic Know Your Customer (eKYC) solutions: false positives in identity verification.

As digital banking, fintech, telecommunications, insurance, and other online services become increasingly popular, businesses depend on automated identity verification to onboard customers securely.

However, an eKYC system that incorrectly identifies legitimate customers as potential fraudsters can create unnecessary friction, damage customer trust, increase operational costs, and ultimately result in lost business opportunities.

Understanding eKYC false positives is therefore essential for organisations looking to balance fraud prevention with a seamless digital customer experience.

1. What Is an eKYC False Positive?

An eKYC false positive occurs when an identity verification system produces an incorrect fraud-related result.

In practical terms, this can mean that a genuine customer is incorrectly flagged for further investigation or rejected because the system interprets their identity information as suspicious.

For example:

• A legitimate customer is flagged because their facial image does not meet the system’s matching threshold.
• A genuine MyKad is incorrectly identified as suspicious due to image quality.
• A customer’s selfie is flagged because of lighting conditions or facial positioning.
• An automated fraud detection system incorrectly associates a legitimate applicant with a suspicious identity profile.

There is an important technical distinction worth understanding.

In general fraud operations, a false positive often means a legitimate customer is incorrectly flagged as fraudulent.

However, in some regulatory eKYC frameworks, including Bank Negara Malaysia’s terminology, a false positive refers to a fraudulent or invalid identity being incorrectly accepted by the verification system.

The incorrect rejection of a genuine customer is generally classified as a false negative in that regulatory framework.

Although the terminology differs, both types of errors demonstrate why identity verification accuracy matters.

Businesses must minimise both incorrect rejections and incorrect approvals.

2. What Causes False Positives in eKYC?

Several technical and operational factors can contribute to inaccurate eKYC decisions.

A. Poor Image Quality

One of the most common challenges involves the quality of identification documents and facial images.

Customers may submit:

• Blurry photographs of their identification cards.
• Images captured in poorly lit environments.
• Overexposed or underexposed selfies.
• Identification documents with reflections or glare.
• Photographs where important information is partially covered.

These conditions can affect Optical Character Recognition (OCR) and facial recognition accuracy.

A genuine identity may consequently receive an unexpected verification result.

B. Facial Recognition Limitations

Facial recognition technology compares facial characteristics between two images to determine whether they belong to the same individual.

However, facial recognition systems are not perfect.

Differences in camera angles, facial expressions, lighting, image resolution, and biometric algorithms can influence matching results.

Research conducted by the US National Institute of Standards and Technology (NIST) also identified demographic differences in the performance of facial recognition algorithms.

This highlights the importance of evaluating biometric systems across different populations rather than relying exclusively on overall accuracy figures.

C. Incorrect Fraud Detection Thresholds

Every automated identity verification system uses decision thresholds to determine whether an application should be approved, rejected, or flagged for further review.

If these thresholds are too strict, legitimate customers may experience unnecessary rejections.

If they are too relaxed, fraudulent applications may successfully pass verification.

The challenge is finding the right balance between security and customer convenience.

D. Document Recognition Issues

Different countries issue identification documents with different layouts, security features, fonts, and information formats.

For example, Malaysian MyKad documents have specific characteristics that may differ significantly from Singaporean NRICs, Indonesian e-KTPs, or international passports.

A verification engine that does not adequately support regional document variations may struggle to extract and validate information correctly.

E. Incomplete or Outdated Verification Technology

Some businesses continue using verification systems that depend heavily on basic OCR or conventional facial matching.

Without additional verification layers, these systems may struggle to distinguish between genuine customers, document manipulation, and sophisticated impersonation attempts.

Modern eKYC solutions should combine multiple verification techniques rather than depend on a single security check.

3. Real-Life Example: How False Positives Affect Fintech Customer Onboarding

False positives are not simply theoretical technical problems. They can have measurable commercial consequences.

A documented case study published by TransUnion describes a UK-based fintech company that experienced significant customer onboarding challenges after introducing additional identity verification controls.

The fintech company implemented document verification and selfie controls to address increasing fraudulent applications involving identity theft and impersonation.

Although these measures helped strengthen fraud prevention, they also introduced friction for legitimate customers.

An assessment revealed:

• 61% false-positive rate: A significant proportion of applications were incorrectly flagged.
• 48% abandonment rate: Many genuine customers abandoned the onboarding process.
• Customer experience challenges: Additional verification requirements made the registration journey less convenient.

The company subsequently introduced a layered fraud prevention strategy using TransUnion’s TruValidate solutions.

Following implementation, the reported results included:

• False-positive rates reduced from 61% to 12.5%.
• Customer friction reduced for 87% of applications.
• Additional fraudulent applications were identified.

This case demonstrates an important lesson: strengthening fraud prevention should not come at the expense of legitimate customers.

An effective eKYC strategy must be capable of identifying suspicious activity while allowing genuine applicants to complete verification with minimal friction.

Source: TransUnion – FinTech Reduces Customer Friction and False Positives While Mitigating Fraud Risks.

4. The Business Impact of eKYC False Positives

For businesses operating digital platforms, inaccurate identity verification can create several challenges.

Customer Abandonment

Customers expect digital onboarding to be fast and convenient.

When genuine applicants repeatedly encounter verification failures, they may abandon the process and move to competing platforms.

This directly affects customer acquisition and conversion rates.

Increased Operational Costs

False alerts create additional work for compliance teams and customer support departments.

Employees may need to manually review identification documents, investigate rejected applications, and respond to customer complaints.

Over time, these activities increase operational expenditure.

Reduced Customer Trust

Identity verification is often the first interaction a customer has with a digital financial service.

Repeated verification failures can create the impression that a business has unreliable technology.

This can negatively affect customer confidence and brand reputation.

Compliance and Fraud Exposure

False positives in the regulatory sense, where fraudulent identities are incorrectly accepted, create another category of risk.

Businesses may unknowingly onboard individuals using forged documents, manipulated identities, or impersonation techniques.

This can expose organisations to financial crime, regulatory penalties, and reputational damage.

The objective should therefore be to improve both fraud detection and legitimate customer acceptance.

5. How Businesses Can Reduce eKYC False Positives

Reducing false positives requires more than simply changing a facial recognition threshold.

Businesses should consider a comprehensive identity verification strategy.

1. Implement Multi-Layer Identity Verification

An effective eKYC process should combine multiple verification methods, including:

• OCR-based identification document extraction.
• Document authenticity and tampering checks.
• Facial matching between the identification document and live selfie.
• Liveness detection.
• Automated identity data validation.

Using multiple verification layers provides a stronger basis for making identity decisions.

2. Improve Image Capture Quality

Businesses should make it easier for customers to submit usable images.

Clear instructions, automatic image quality checks, and real-time capture guidance can help reduce errors caused by poor lighting, blur, reflections, and incorrect positioning.

3. Use Intelligent Risk-Based Decisions

Not every verification failure should result in an immediate rejection.

Where appropriate, businesses should establish different decision paths:

• Low risk: Automatically approve applications that satisfy verification requirements.
• Medium risk: Request additional verification or route the application for manual review.
• High risk: Reject applications with sufficiently strong indicators of fraud.

This approach helps avoid treating every technical irregularity as evidence of fraudulent activity.

4. Continuously Monitor Verification Performance

Businesses should regularly evaluate their eKYC systems using measurable indicators such as:

• False Acceptance Rate (FAR).
• False Rejection Rate (FRR).
• Facial Matching Error Rates.
• Manual Review Rate.
• Customer Onboarding Completion Rate.
• Average Verification Time.

Monitoring these indicators helps organisations identify weaknesses and improve their verification processes.

5. Select an eKYC Provider That Understands Regional Identity Documents

For businesses operating in Malaysia and across Southeast Asia, regional document support is particularly important.

An effective verification solution should understand local identification formats and account for variations in document layouts, languages, and security features.

This can help improve the consistency of identity verification across different customer groups.

6. The Importance of Balancing Fraud Prevention and Customer Experience

One of the biggest misconceptions about digital identity verification is that stricter security automatically means better protection.

In reality, an excessively restrictive verification process can create unnecessary customer friction without necessarily delivering better fraud detection.

Similarly, a system that prioritises fast approvals without sufficient verification can expose businesses to identity fraud.

The ideal approach is to establish a balanced eKYC framework that:

1. Accurately identifies legitimate customers.
2. Detects forged or manipulated identification documents.
3. Verifies that the person submitting an application is physically present.
4. Identifies potential impersonation and spoofing attempts.
5. Provides actionable information when verification fails.
6. Supports manual investigation where automated decisions require additional scrutiny.

Businesses should view eKYC as an ongoing risk management capability rather than a one-time identity checking exercise.

7. How LangitDigital’s Epsilon ID V3 Helps Address eKYC Verification Challenges

Reducing identity verification errors requires technology that can assess multiple aspects of an applicant’s identity rather than relying on a single verification method.

This is where LangitDigital’s Epsilon ID V3 comes in.

Designed for businesses that require reliable digital identity verification and fraud prevention, Epsilon ID V3 combines several verification capabilities within a single eKYC workflow.

Its capabilities include:

AI-Powered ID Document Verification

Epsilon uses Optical Character Recognition (OCR) to extract identification information and validate documents, including Malaysian MyKad and supported international identification documents.

It also checks for document tampering and digital alterations.

Facial Matching

The platform compares a customer’s live selfie against their identification document photograph to help verify that both belong to the same individual.

Passive Liveness Detection

Epsilon incorporates passive 3D liveness detection to identify presentation attacks involving printed photographs, screens, physical masks, and AI-generated deepfakes.

Intelligent Data Validation

Extracted information is cross-checked against relevant document standards, including ICAO-related validation rules such as Machine Readable Zone (MRZ) checksums and document expiry dates.

Actionable Verification Results

When an application fails verification, the system can return information about the failure reason through its webhook integration.

Businesses can use this information to determine whether an application should be rejected, investigated, or routed for manual review.

With verification designed to complete in under 10 seconds, Epsilon ID V3 aims to help businesses strengthen identity verification without unnecessarily slowing down customer onboarding.

Importantly, no eKYC solution can guarantee zero false positives or false rejections. Businesses should continuously monitor their verification performance and maintain appropriate human review procedures for exceptional cases.

For organisations looking to improve their digital onboarding processes, Epsilon provides an integrated approach to identity verification and fraud prevention.

Discover Epsilon ID V3:

https://langitdigital.my/products/epsilon

Conclusion: Better eKYC Means Better Security and Happier Customers

eKYC false positives remain an important challenge for businesses embracing digital customer onboarding.

Whether caused by poor image quality, facial recognition limitations, document recognition errors, or inappropriate verification thresholds, inaccurate decisions can affect both security and customer experience.

The solution is not simply to make verification stricter. It is to make verification smarter.

By implementing multi-layer identity checks, intelligent risk assessment, continuous performance monitoring, and appropriate manual review procedures, businesses can create a more reliable digital onboarding experience.

With solutions such as LangitDigital’s Epsilon ID V3, organisations can take practical steps towards strengthening identity verification, reducing unnecessary friction, and protecting their digital platforms against increasingly sophisticated identity fraud.

The goal is simple: Verify genuine customers confidently. Identify suspicious identities accurately. Make digital onboarding easier for everyone.