Managing Bank Fraud and High-Risk Customers: A-Z Guide

Modern banks handle a variety of financial transactions. They deal with a huge number of money flows and enormous money volumes. It is no surprise that banks have become prime targets of fraudulent activities. Impressive $10bn loss was reported because of fraud in 2023 with nearly 1 in 5 people affected by the imposter scam.

A serious problem for banks is that modern criminals are highly tech-savvy. They actively use the latest technologies to access user data and money. Considering this, financial organizations must look for advanced methods to prevent banking fraud.

What are the methods to prevent bank fraud? How are IT technologies involved in fraud prevention? How do financial institutions work with high-risk customers?

In this article, I will answer these questions. Based on my expertise in fintech and banking development, I’ll share the most advanced methods of bank fraud protection.

What Is Banking Fraud and its Examples?

Banking fraud is a fraudulent activity used to obtain unauthorized money.

What is bank frauds examples?

Bank fraud examples include:

• Loan fraud is obtaining a loan by using someone else’s personal data. With the help of loan fraud, the attacker receives the required amount of money from the bank. At the same time, the person who actually owns the personal data is left with a debt
• Card fraud is unauthorized use of someone else’s card information for malicious purposes. Fraudsters can obtain card data by stealing the physical card, skimming the card data, or deceiving the cardholder into sharing the financial information voluntarily.
• Internet fraud includes phishing, vishing, and other identity theft schemes. As a part of Internet fraud, attackers can pretend to be bank employees and force the customer to provide their login credentials. They can then use these credentials to log into the  customers’ bank accounts and use the customer’s money for their own purposes.
• Occupational fraud is fraud committed by company employees using their official position. For example, an employee might create fake invoices for goods or services that were never provided, pocketing the money for themselves, or they might create fake bank statements and forge signatures to embezzle funds from company accounts.

How Banks Execute Protection from Fraud

Banks enable fraud protection by

• Screening customers before entering into relationships with them
• Monitoring monetary transactions as they are executed

Each fraud prevention technique is essential for establishing robust data security and eliminating criminal activity. However, it is essential to know the pros and cons of customer screening and transaction monitoring to develop a comprehensive system of protection from fraud.

Customer screening

Pros:

• Immediately indicates if there are any risk-related issues with the customer.
• Provides a general understanding of customer type and lets bank adjust the type of treatment for the customer to increase security

Both participants of every transaction are screened, which ensures reliable funds protection

Cons:

• It is a slow process, as it results in additional investigations performed by security/compliance officers rather than immediate action.

Example of customer screening in a bank

Let’s review an example of customer screening based on the Dow Jones list, which we usually use in our fraud prevention software projects.

When a customer’s information is checked against the Dow Jones sanctions list, the system generates alerts for even the smallest data matches. The software analyzes all available customer data, such as name, surname, family members, passport information, and employment details. After the screening, the system may produce zero or more matches, along with a match percentage (e.g., 70%). T

The percentage is determined by the number of data matches and their respective significance. For example, a passport data match would have a higher value than a date of birth match. Potential negative results: Sanctioned, PEP.

Now, let’s transit to the example.

Note: the example is not a real person and is only for explanation purposes. Only a small part of the evaluated data is showcased.

Imagine that we have a customer with the following data

1. Name: John Doe
2. Date of birth 01.01.1991
3. Passport ID: 1234567890
4. Job: Manager
5. Nationality: Philippines
6. Country of residence: Philippines

Let’s assume there is the following sanctioned individual in the Dow Jones list:

1. Name: Richard Smith
2. Date of birth: 01.01.1991
3. Passport ID: 1234567890
4. Job: Unemployed
5. Nationality: Mozambique
6. Country of residence: Philippines

There is a high chance to detect the match because the date of birth, country of residence, and passport ID all match, despite differences in name, job, and nationality.

The scheme below demonstrates the customer screening in a banking fraud protect software based on the Dow Jones list:

Transaction monitoring

Pros:

• Analyses the user’s behavior and suggests outcomes for any detected suspicious actions
• Adapts quickly to changes in customer behavior
• The decision-making process is very fast, as it is fully automated

Cons:

• Most efficient only when enough data is available, which requires a robust data architecture of the bank

Example of transaction monitoring in a bank

I will explain how the transaction monitoring is executed based on our recent project of implementing a fraud management system. I want to point out that I cannot disclose the calculation rules of this fraud management system. However, I can mention that those rules are mostly built based on customer behavior and identify suspicious behavior patterns. The possible outcomes of transaction screening are to approve or reject.

So, let’s assume the customer is an active user of the bank for 6 months now and the fraud management system has already gathered enough data to analyze the behavior of the customer. For example:

• Their income is $5000 per month
• Their monthly expenses are $4500
• Their mortgage payments are $1000 per month
• They spend $1500 per month on loan payments
• They sometimes send $300 to the account of their spouse
• They never make any payments to other accounts

The data above means the customer spends almost all their income every month, leaving them with approximately $500 in savings. Most of the customer’s expenses are mandatory payments.

Let’s assume the customer saved $1000. One day during the night time, four payments to four different new accounts are made, $200 each.

This behavior doesn’t look typical for the customer, right? So the transactions are getting blocked and require either approval or additional clarification from the customer before being processed.

As a result, possibly fraudulent transactions (made by a presumably hacker) were prevented, and the customer’s money was saved. Any illegal activities or potential money laundering payments were also prevented, depending on the outcome of the investigation of the transaction nature.

The scheme below demonstrates the transaction monitoring in an fraud management system:

What Fraud Prevention Checks Do Banks Conduct?

Banks conduct different types of fraud prevention checks at different stages of bank-customer interaction. They are customer onboarding check, monetary transaction check, non-monetary activity check, and overnight check. Below, I review the fraud prevention checks in more detail.

Customer onboarding check

At the onboarding stage, the bank initiates relations with the customer and performs the Know Your Customer (KYC) checks. Here is how the KYC fraud protection is executed:

Data acquisition. The customer provides its data either in the bank branch or online. They usually upload the copies of Personal ID and Tax ID into the banking system and complete the KYC questionnaire

Initial data screening. The customer data is screened using a fraud evaluation software system. This system checks customer background, including loan history, criminal records, digital footprint and indicates if any suspicious activities were or may be performed by the user.

The second-round data screening. Bank fraud prevention system screens customer data to identify if they are in the sanctions list or qualified as a politically exposed person (PEP). If any matches are found, customer onboarding is put on hold until further investigation is done.

Risk rating assignment. The bank assigns the risk category to the customer. The customers are usually categorized as low risk, mid risk, and high risk customers. The risk assessment logic is implemented based on the National Bank guidelines which vary by country.

Monetary transaction check

The procedure of fraud prevention in banks is executed every time the monetary transaction is performed. This applies to both incoming and outgoing transactions. Here is how banks protect from fraud during money transfers:

Sender and recipient screening. This step is similar to second-round data screening performed at the onboarding stage. The transaction participants are checked for sanctions and the presence in the REP lists. If a REP affiliation is detected, the fraud prevention mechanism suspends the transaction.

Transaction monitoring. Even after the payer and payee have successfully passed the fraud prevention check, the bank does not stop monitoring the transaction. During the process of money transfer, the anti-fraud system analyzes transaction patterns and flags any suspicious activity. For these purposes, machine learning (ML) and artificial intelligence (AI) in baking technologies are often used. They help quickly recognize suspicious activities that may be invisible to humans. If the transaction is qualified as suspicious, it is put on hold.

Non-monetary activity check

User activity monitoring. Every action a user performs in a bank branch, web app or mobile app is monitored and analyzed by the fraud protection software. If action performed is qualified as suspicious, the user account is temporarily restricted until further investigation is conducted.

For example, the bank can check how frequently a user logs into their account, the locations from which they log in, and the devices they use. If there is an unusual login pattern, for instance, multiple logins from different countries within a short time, the system can flag this behavior as suspicious and trigger additional verification steps to ensure the account fraud security.

Personal data change monitoring. If the user changes the information specified in the personal ID, for example, first or last name, this is a reason to repeat the procedure of prevention of bank frauds. In this case, all steps specified in the Onboarding section are activated. That is, first, customer documents are collected, then initial and second-round screenings are carried out, and then the customer is assigned a risk category.

Overnight check

Every night, the bank performs the same fraud protect procedures as during customer onboarding, except those are executed for all existing customers. This is necessary in order to promptly identify fraudulent activities that occurred after the last check. Regular checking ensures that customers are placed in the correct risk category and significantly reduces the likelihood of fraud. As a result, data and transactions are reliably protected.

Dealing With Low-, Middle, And High-Risk Customers

One of the main activities in the fraud prevention in banks is the categorization of customers into risk categories: low, middle, and high. For each category, the bank then applies special actions aimed at ensuring data safety and preventing malicious activity.

Here is what customers fall under the specified risk categories and how the bank executes a fraud protect strategy depending on the customer risk category.

Low risk

Low-risk customers are customers whose sources of income can be easily identified and whose transactions are executed within the established bank limits.

Typical representatives:

• Institutions that regularly make their financial statements publicly available
• Government institutions
• Financial organizations licensed by central bank
• Employees whose salary is clearly defined when opening a bank account
• Self-employed individuals whose income statements confirm the carried out transactions 

How to prevent bank fraud:

• Standard monitoring: routine transaction monitoring and regular fraud checks
• Basic KYC procedures: initial onboarding checks without the need for extensive documentation or frequent reviews
• Limited restrictions: few restrictions on transactions and account activities, given the low likelihood of fraudulent behavior
• Standard alerts: normal alerts and notifications for unusual activities

Middle risk

Middle-risk customers are those whose sources of income are not as transparent or consistent as low-risk customers. They may have irregular income patterns or engage in transactions that are slightly outside the typical limits set by the bank. 

Typical representatives:

• Non-regulated companies with up-to-date tax clearance
• Small businesses and startups
• Freelancers and consultants with irregular income
• Individuals with multiple income streams or investments
• Customers with moderate transaction volumes that occasionally exceed standard limits

How to prevent bank fraud:

• Enhanced monitoring: more frequent transaction monitoring and periodic reviews to identify any suspicious activity. This involves automated systems that flag irregular transaction patterns for further investigation
• Additional KYC procedures: require extra verification steps during onboarding and at regular intervals. This could include submitting updated documentation, undergoing additional identity verification processes, or answering security questions
• Moderate restrictions: some restrictions are placed on large or unusual transactions until further verification is completed. This could mean temporarily holding transactions that are out of character for the customer until they are manually reviewed and approved
• Increased alerts: more frequent alerts and notifications for activities that deviate from normal patterns. This ensures that both the bank and the customer are promptly informed of any potential fraudulent activities.

High-risk

High-risk customers are those with complex, high-volume transactions, inconsistent income sources, or those previously flagged for suspicious activities. High-risk customers often require the most stringent fraud prevention measures.

Typical representatives:

• High-net-worth persons with a wide range of investments
• Companies working in high-risk industries, including gambling, cross-border operations, car dealership, pawn shops, etc.
• Companies providing private investments or acting as financial intermediaries
• Customers who have a history of suspicious activity
• Politically exposed persons 
• Customers from high-risk countries

How to prevent bank fraud:

• Intensive monitoring: using advanced bank frauds detection systems to scan every transaction and flag suspicious activities for immediate review
• Comprehensive KYC procedures: extensive due diligence during onboarding, including in-depth background checks, verification of multiple identity documents, and, in some cases, biometric verification.
• Stringent restrictions: imposing strict limits on transaction amounts and types until thorough verification is completed. Large or unusual transactions are routinely held for manual review before approval.
• Frequent account reviews: quarterly or even monthly reassessments of the customer’s risk category. This ensures that any suspicious changes are promptly addressed.
• Enhanced alerts: real-time notifications to both the bank’s fraud prevention team and the customer

What Actions Are Taken if Banking Fraud is Suspected?

Whenever further investigation is necessary, the security/compliance officers must gather additional details from the customer to assess the customer’s risk level. These details may include:

• Proof of income. The security officer may request recent pay stubs, tax returns, or bank statements that show the income deposits.
• Source of funds. This must contain evidence of how the customer obtained significant sums of money. The examples of such evidence may be property sale records, inheritance documents, or investment portfolio statements.
• Transaction history. A comprehensive review of the customer’s transaction history to identify any unusual or high-value transactions that deviate from their normal behavior.

They also need to obtain approvals from managers, the Head of Security/Compliance, and sometimes even the Chief Compliance Officer (CCO) and Chief Executive Officer (CEO). Based on these investigations, the outcome could be to either proceed with onboarding/dealing with the customer or to stop onboarding/dealing, which would result in the account closure process.

How Technologies Enhance Fraud Protection in Banks

Today, no financial institution can do without fraud prevention software

Such software

• Evaluates each customer and transaction 
• Classifies customers and transactions by risk level
• Blocks non-typical operations
• Creates customer behavior models
• Simplifies the payment process for trusted customers
• Sends an additional transaction confirmation request to risky customers
• Enables other fraud-prevention measures

Artificial intelligence (AI) plays a crucial role in fraud prevention in banks. Using AI, developers create algorithms that quickly and accurately analyze big data, starting with those on the customer device and ending with credit card data.

AI-based anti-fraud software provides banks with unlimited flexibility in fraud management. It allows organizations to conduct unbiased evaluation, provide instant response to malicious activity, and minimize a chance of error.

To develop fraud prevention software, you need a team with extensive experience in fintech development and comprehensive understanding of the regular basis in this industry.

At DashDev, we possess the skills mentioned above and invite you to discuss your anti-fraud software project with us.

Contact us today!