Legacy infrastructure still hindering bank AI capabilities

From risk management to fraud detection, banks would need to move from legacy infrastructure to fully embrace the capabilities of AI.

The financial industry is often regarded as the first industry to adapt to new technologies. While there are still regulatory requirements that need to be adhered to when developing and deploying new technologies, the reality is, if a use case is successful in the financial services industry, it will most likely be successful in other industries as well.

Take GenAI for example. According to a report by IDC commissioned by SAS, APAC organizations are rushing to jump onto the AI bandwagon, with nearly half (43%) planning a large investment increase in AI of over 20% in the next 12 months.

While banks have been using AI for some time for risk management, fraud detection, customer service and even to offer personalized financial experiences, the use of GenAI applications is still not at a level that it should be. The reason for this is simply because banks are often hindered by legacy infrastructure that makes it harder for them to pivot adopting new technologies.

For SAS, enabling banks to work with technologies is what they do best. The data analytics company has had a strong presence in the Southeast Asian region, enabling banks to make the most of their data to develop more use cases while adhering to regulatory and compliance requirements.

In Malaysia, SAS worked with Alliance Bank to implement Asset Liability Management (ALM) to bolster its risk monitoring and oversight for its interest rate risk and liquidity risk management. By enabling advanced stress testing, scenario analysis, and predictive analytics capabilities, the bank is able to make faster and more accurate business decisions in response to a changing interest rate environment and gain competitiveness.

When it comes to AML, a technology study from SAS and KPMG revealed that while adoption of AI and machine learning remains modest, interest in GenAI technology is robust but industries are still cautious. Nearly half of respondents say they are currently piloting GenAI (10%) or are in the discovery phase (35%).

According to Ahmed Drissi, APAC AML lead for fraud and security intelligence, the use of AI and machine learning in anti-fraud programs is expected to nearly triple over the next two years. Three in five organizations also intend to increase their budget for anti-fraud technology over the next two years.

On addressing concerns of false positives, Drissi pointed out that SAS continues to improve its data capabilities and has been able to reduce both false positives and negatives.

In order to be successful in this deployment, data management is key. As banks face a veritable data explosion from multiple sources, effective data management and data governance frameworks have never been more essential. In an SAS survey, improved risk management (64%), improved customer experience (55%) and improved fraud detection (51%) are seen as top benefits for consolidating customer data. Yet only 14% intend to significantly consolidate customer data, and fewer than half (44%) say the same for non-customer data.

This is where SAS has invested heavily in SAS Viya. The cloud native data and AI platform allows users to manage, analyze and transform data into actionable insights. Specifically, it's designed for scalability and can handle large datasets, offering tools for data preparation, model building, and deployment.

Despite the capabilities of SAS Viya, incumbent banks would still need to modernize their legacy infrastructure in order to get the best value in AI. Febrianto Siboro, Managing Director for SAS in Malaysia, Indonesia, Philippines and Vietnam, mentioned that they are currently working with major banks in the region on this, especially as the incumbent banks look to deal with challenges from the rise of fintech and digital banks.

For Naeem Siddiqi, Senior Advisor for Risk and Quantitative Solutions at SAS, once banks are able to move beyond the challenges in their legacy infrastructure, they will have access to greater opportunities, especially in developing products and services for customers.

One area in particular is developing products to improve financial inclusivity and access to financial services. Digital banks are successful in this area simply because they lack legacy infrastructure and are also capable of leveraging non-financial data to offer financial products. For incumbent banks, this is no easy task, but it does not mean that they will not be exploring these capabilities as well.