Artificial Intelligence and ISO 20022 Improve Payment Transactions

Artificial Intelligence and ISO 20022 Improve Payment Transactions

Written by Niranjan Govindaram

Modern payment systems are imperative for any country that wants to grow and prosper in a global marketplace. Efficient and secure payment systems are the central point of a country’s economy as they drive its business and enable its overall economic activity. Reliable payment systems cater to different segments of the population and companies.

As seen in developing countries like India, payment systems also play a critical role in financial inclusion. This allows greater access to government benefits directly and more transparency without the need to go through intermediaries or deal with the bureaucracy.

Evolution of payment platforms and transactions

Payment processing has dramatically changed over the past several decades, evolving from cash-based to digital payment methods. In the age of cash-based transactions, security risks were limited. Technology has made payment systems efficient and transparent, but the risks have increased exponentially.

Payment systems have come a long way, from the days of checks to electronic platform-based batch processing platforms to the current instant payment platforms. Several decades ago, when checks were the primary payment transaction instrument, processing payments could take more than five days. Today, global instant payment systems, like Fedwire, TARGET2, RTP, TIPS, RT1, and FedNow, process transactions in only a few seconds. These new methods have expanded business and consumer transactions immensely, making it easier, faster, and more secure to make and receive payments. Almost every country has an instant payment platform or is in the process of implementing a platform.

Current payment transaction challenges

Even though payment transactions are processed in seconds, there are some challenges surrounding digital payment platforms. Common challenges include:

1. Online fraud and security. As more people use digital payments worldwide, there are increased fraudulent transactions with the enormous payment transaction volumes. According to a 2021 report from Juniper Research, payment fraud costs billions of dollars for users. It is important to recognize the different types of fraud that users fall prey to and create strategies to manage and mitigate each type.
2. Integrated payments. The exponential growth of digital payments has also seen a rapid rise in customer expectations, especially corporate customers. A corporate entity’s digital transaction capabilities include accommodating multiple vendors and enhancing payment options for a smooth and seamless user experience. As such, it is essential for financial institutions to offer enhanced corporate services like cash management, which includes managing payments, collections, liquidity, proper payment integration, and faster adoption of new payment methods. The complex technology required to integrate a corporate entity’s platform with a bank’s platform has several components, so the choice of provider can have a significant impact. With multiple payment methods now available, corporate customers look to financial institutions to make intelligent decisions about which payment method to use.
3. Cross-border payments. These transactions involve a payer and beneficiary in different countries or regions. Technology has improved the speed and security of transactions, and the volume of cross-border payments is growing rapidly. This also means more payment transactions are vulnerable and exposed to fraud risks.
4. Customer service. Customer expectations have increased with the rapid usage of technology in payment initiation and processing. Online payment processes are more complicated and riskier. Technology empowers customers to expect quick and extensive support. Customers do not want to feel restricted by their technology; they want instant customer service via their tablet, smartphone, or other interface. The evolution of payment ecosystems means that consumers expect the same level of support from their smartphone as on a desktop.
5. Transaction data. Data quality and unstructured data are critical issues in payment transactions. Since most banks have different message formats across their payment transaction lifecycle, converting the data from one message format to another means a loss of data while processing each step of a payment transaction. This leads to a significant dependence on human interaction to interpret, repair, and process the payments.
6. Manual intervention. Multiple message systems coupled with legacy systems across the payment processing lifecycle result in the failure of payments, which are then moved to the manual repair queue for further investigation and processing. This problem is more acute in cross-border payments processing. It results in potentially lengthy payment processing delays for end users and contributes to the cost of payment processing for banks, impacting the overall client experience.

ISO 20022 and its impact on payments

ISO 20022 messaging standard is a unified and single approach used in all financial messaging, including payments, card services, and foreign exchange. It applies to domestic, cross-border, automated clearing house (ACH), and instant low-value real-time and high-value real-time payments. A common messaging standard across all regions and countries means better communication, richer and more accurate data, reduced risks, and improved traceability. Interoperability between payment systems within and between countries will improve as ISO 20022 follows XML-based approaches.

Financial institutions are migrating their payment systems from message type (MT) and other legacy and in-house messaging standards to the more structured ISO 20022 messaging standard, which means they will all use the same message framework and meaning. By the end of 2025, most payment systems worldwide are expected to move to the ISO 20022 messaging standard.

In addition, ISO 20022 enables the transfer of important and detailed information for any payment transaction, such as the purpose of the payment and the structured information of the sender and recipient, including their addresses. This level of detail makes it easier for financial institutions to identify potential risks associated with a transaction and take appropriate action to prevent sophisticated financial crime. Even though a leap to the new standard means logistical and technical obstacles to ensure success, this move helps businesses streamline operations, ensure “greater traceability, reduce fraud and human error, and enhance cross-border trade.”

How artificial intelligences help mitigate payment challenges

Machine learning (ML) is an essential subset of artificial intelligence (AI). It involves the use of algorithms that can learn from and perform predictive data analysis. Predictive analytics employ “statistical algorithms and ML techniques to identify the likelihood of future outcomes based on historical data.” In payments, these algorithms are typically used for fraud detection and enable forecasting transaction volumes, which is crucial for managing resources and ensuring seamless operations, especially during peak sales periods, predicting consumer spending behavior, and optimizing transaction processes.

Neural networks are a type of AI influenced by the human brain’s functioning that can help identify patterns, abnormalities, and inconsistencies in large datasets. They are instrumental in fraud detection and identify unusual transactions that deviate from established patterns based on the history of a particular user.

The positive effects of using AI and ISO 20022 for payment processing

With the expansion of structured data fields, the ISO 20022 message standard provides financial institutions with better opportunities to benefit from the efficiency and performance benefits of deploying AI functions, including ML algorithms, to automate payment screening and processing. With the standardization of data using ISO 20022, these algorithms can analyze transactions in real time and reduce the risk of evading illegal activity detection. This capability improves the speed and accuracy of the screening process, reduces the cost of manual operational review, and decreases the risk of human error.

An intelligent and innovative screening approach is best described as a grid where payment transaction data elements and tags appear in columns, while the information it can be matched against is in rows. Each cell can then provide the recommended and appropriate screening behavior for each payment transaction. This targeted approach ensures the correct processing of all payment transaction data, allowing financial institutions to ensure compliance and detect fraud while avoiding false positives linked to mismatches between information types (e.g., creditor name does not match against street name).

The combination of AI and ISO 20022 for payment transactions is not just a trend but a significant step towards a more secure, efficient, and user-friendly unified payment ecosystem. By understanding and leveraging the benefits of AI and ISO 20022, payment industry players, including banks and financial institutions, can improve operations and contribute to shaping a contemporary, advanced, and innovative payment landscape.

About the Author:

Niranjan Govindaram is a corporate banking subject matter expert with specific expertise in payments and cash management. Niranjan holds a master’s degree in business administration with a specialization in strategy and management from Queens University, Kingston, Canada. He can be reached at niranjn@gmail.com