Back to Blog

KYC Automation In Banks Using Computer Vision

computer vision in banking

Digital transformation is redefining the way industries operate. Every industry, including banking, is focusing on a customer-centric approach. In order to meet customer expectations, banks are adopting advanced technologies to provide seamless mobile banking and digital payment services. With the rise of online transactions, there have been growing incidents of phishing attacks and frauds. 

Hence, banks follow a KYC or ‘Know Your Customer’ process to verify customer identities and their financial dealings. Here, computer vision technology can be used to enhance the process of customer verification. In this blog, I discuss the current KYC process in banks and how computer vision can be used to automate it.

Table of Contents

What is KYC - An Introduction

The term ‘KYC’ stands for Know Your Customer or Know Your Client. The KYC process helps banks and other commercial lending institutions to understand their customer’s intentions by carrying out due diligence. The main objective of this process is the prevention of money laundering and management of financial risk and potential frauds. KYC forms the core of every banking process and involves 3 key elements:

1) Customer Identification Program or CIP

In the US, it is mandatory that a customer's identity needs to be verified before indulging in any financial transaction. It has to be identified whether the customer is an individual, sole proprietorship, partnership, company, or LLP. This CIP initiative is intended to prohibit any illegal activities, like money laundering,  corruption, or terrorism funding. The minimum criteria for opening a new bank account include - name, date of birth, address, and unique identification number.

2) Customer Due Diligence or CDD

This step ensures the trustworthiness of a potential customer. CDD is a key element for risk management as it examines whether the person is a criminal, PEP (politically exposed person) or a terrorist. In this process, a potential customer is categorized into different risk levels based on the following levels of due diligence:

  • Simplified Due Diligence - In this case, the risk for terror funding or money laundering is low.

  • Basic Customer Due Diligence - Here, information is obtained from all customers in general for identity verification and risk evaluation.

  • Enhanced Due Diligence - At this level, additional information is gathered for customers with associated high risks to get an in-depth understanding of the customer’s financial activity.

Related: Download the checklist that lays out everything you should consider  before implementing a machine learning project or workflow →

3) Continuous Monitoring

The financial transactions and accounts of customers need to be monitored on an ongoing basis. Depending on the customer’s profile and risk mitigation strategy, few other things should also be monitored - unusual cross-border transactions, a sudden spike in activity, unfavorable media mentions, etc.

Who needs KYC?

In the wake of recent money laundering, fraud, and tax evasion cases, global attention is focused on AML(anti-money laundering) and CTF (counter-terrorism financing) directives issued by financial institutions. Hence any professional or business which falls under the ‘regulated sector’ has to compulsorily adhere to the KYC regulations. These include:

  • Banks

  • Credit institutions

  • Legal professional

  • Real estate agents

  • Tax accountants and financial advisors

So these professionals and businesses usually follow the traditional KYC process, which I have discussed below.

The current KYC process

The traditional KYC process requires a customer’s ID proof and address proof. The customer can present his passport, driving license, voter ID card, or PAN card as a valid identity proof. Documents like electricity bill, landline bill, bank passbook, rental agreement or passport copy can act as proof of address. Presently, the KYC Process is done through the following 2 channels -

Online - By uploading ID and address proof documents in an online submission form. These documents are manually verified depending on the quality of images. If the documents are readable, the bank representatives get in touch with the customer to complete the process.

Offline - Here, the customer has to download the KYC application form. Take a print out and fill in the details and submit it physically to the respective authorities with attested copies of specified documents.

To get rid of the lacuna in the current process, computer vision can be implemented to make the future KYC process more efficient and accurate.

Future KYC process: Automation using computer vision

Computer vision technology is used to analyze and extract relevant information from images. It’s been used in banks to streamline and automate the KYC process. Below, I discuss a few amazing practical applications of computer vision in banking. This will help you in determining how AI could be a perfect solution for any bank or financial institution.

Simplify customer verification process

When opening new accounts banks need to verify the customer's identity and assess any associated transactional risks. The usual KYC (Know Your Customer) process is time-consuming but the use of computer vision can reduce it to a few minutes.

Many banks are already using computer vision technology to make the process convenient. A quick example is BBVA (Banco Bilbao Vizcaya Argentaria) bank, which uses computer vision for digital KYC verification. Customers can complete their digital verification process through a quick video call sitting at home. It’s beneficial for banks as well as they can attract more customers without spending too much effort or time.

Secure transaction using biometric iris scanning

Iris scanning

In recent times cashless transactions have gained momentum as it limits chances of tax evasion and money laundering. Cashless payments are convenient for the consumers but carry the risk of identity and card theft. Iris recognition is by far considered the most reliable payment system due to its stability and accuracy. This biometric technology confirms a person’s identity by capturing high resolution and clear pictures of the human iris. It helps banks in protecting sensitive information by preventing any security breach or unauthorized access.

Good examples of successful biometric implementation are launch of MasterCard's ‘Selfie Pay’ and Visa Multi-factor biometric authentication.

Fraud prevention

The most prominent case of banking fraud till date is the scary ‘Panama papers’ scandal.The Panama papers case deals with leak of 11.5 million files from database of Mossack Fonseca, world’s 4th largest offshore law firm. These files were shared by a German media agency to the International Consortium of Investigative Journalists (ICIJ). The documents reveal how rich and influential personalities, including senior national leaders have been exploiting offshore tax havens. The Panama-based law firm Mossack Fonseca has operations across 42 countries and acts as a registered agent for over 2,00,000 companies involved in illegal trade.

Post the ‘Panama Papers leak’ case, severe concerns were raised about compliance issues and the vulnerability of banking operations.  Banks are now following stringent KYC implementation guidelines for authenticating customer identity and documentation. Banks are now expected to monitor transactions that are considered suspicious to prevent any instances of fraud and money laundering. If any transaction is conducted by professional intermediaries like Chartered Accountants, legal firms, or stockbrokers, banks need to identify the ultimate beneficiaries.

Computer vision can be used to detect any cheque tampering, by identifying missing or duplicate cheque numbers. Recently Citibank launched an image-based interface called Total Wealth Advisor (TWA). This is an easy to use platform containing details about the client’s investments and financial goals. Here, computer vision is being used to stay updated with the client’s transaction records and asset management.

Paving the way for a secure future

The banking domain has witnessed several ground-breaking initiatives in the last few years.  A few noteworthy changes include mobile banking, real-time money transfer services, and e-banking. But with all the good changes, there have been few negatives as well, like the rise of cyber crime, online frauds, and illegal money transfers.  Venturing into the world of AI and computer vision, banks can ensure complete transparency in terms of KYC implementation. They can carry out seamless customer identification and authentication. Banks will also benefit through increased operational efficiency, improved employee productivity, and save time as well as costs.

If you want to integrate AI into your banking operations, can help you achieve it.  Try out no code platform and usher into a new era of digital banking.