AI Adoption Lag Could Leave Asian Financial Institutions Vulnerable

New research from SymphonyAI and Regulation Asia reveals legacy systems, data quality, model explainability, data privacy, and regulatory uncertainty hinder AI adoption in financial crime compliance.

As generative AI adoption continues to increase in Asia Pacific, there are concerns that organizations that have yet to begin their AI journey will be left behind. These organizations are still worried about how generative AI could impact their data or are still not sure where exactly to begin implementing AI solutions.

In Southeast Asia, there are fears of a digital divide in some industries with businesses that have better budget allocation and a skilled workforce taking advantage of the technology.

Interestingly, one industry that continues to see increased adoption of generative AI is the financial services industry. Banks, insurance companies and other financial institutions deal with a lot more sensitive data compared to other industries and would need to meet all regulatory requirements when implementing new technologies.

While there was initial concern on how generative AI could impact data security and privacy in the industry, more financial institutions have begun to build their own AI applications to improve their product offerings and understand customers better.

Despite this, a report by Symphony AI with Regulation Asia revealed that the increased adoption of AI among financial institutions (FIs) in Asia has led to a gap in the industry, with some of them vulnerable to escalating financial crime. Titled “Untapped Potential: AI-enabled Financial Crime Compliance Transformation in Asia – Maturity, Applications, and Trends,” the report is based on surveys and interviews with 126 financial crime compliance, operational, and technology practitioners from FIs across the Asia Pacific region.

Apart from detailing the lack of AI adoption by Asian FIs despite clear benefits and cost advantages in effective financial crime prevention and detection, the results also revealed that over 50% of FIs in the region are not currently using AI for anti-money laundering (AML).

Even before generative AI became mainstream, FIs in Asia have long focused on using AI for AML as it not only speeds the process but has also reduced concerns on false positives that often happen when FIs initially started using AI in AML.

Money laundering risks in Southeast Asia remain high with another report from Moody’s revealing money laundering risks events have climbed 64% in 2023 from 2018. The top five countries that involved high risks on money laundering include Thailand, Singapore, Malaysia, Indonesia and the Philippines.

Using AI to mitigate financial risks

Money laundering is just one of the many financial risks FIs are facing today. Fraud remains the biggest issue in the region. However, FIs have invested heavily in fraud detection capabilities to deal with the issue, with some even focusing on generative AI for fraud detection. In fact, AI-based transaction monitoring, sanctions screening, and fraud prevention have delivered proven benefits as adoption gains traction.

According to the report, while interest in AI is high, only 15% of FIs in Asia say they are actively applying the technology for AML processes. One of the core reasons for this is because many firms are limited by AI integration and data quality. The report stated that integrating AI with existing systems, data quality and availability , model explainability, and data privacy and protection were among the top challenges cited by respondents.

Differing regulatory standards across different markets are also one of the reasons for the slowdown in using generative AI to mitigate financial risks. 58.6% of respondents also cited challenges with legacy systems and data quality as major roadblocks to AI adoption.

Many FIs still see AI as a long-term project, especially the perceived complexity of integrating or overlaying AI into legacy systems. This struggle to effectively implement AI is particularly concerning given the rapidly evolving nature of financial crime. As criminal activity becomes increasingly sophisticated and transcends borders, traditional compliance methods are proving woefully inadequate.

At the same time, the success of commercial generative AI use cases has also witnessed boards and senior managers playing a critical role in driving AI adoption with 40% of respondents saying their top leaders are primary advocates. However, the demonstrable value of AI through reducing false positives, improving accuracy and efficiency, and controlling costs is crucial for board-level AI investment buy-in.

But it's not all distressing news as the report revealed that FIs see AI as an essential solution for effective transaction monitoring. 78% of respondents stated it is a top priority area for deployment especially in the technology’s capabilities to efficiently process vast amounts of data to detect suspicious patterns that traditional methods might miss.

Delivering value from AI

For Gerard O’Reilly, managing director of APAC, Financial Services at SymphonyAI. Financial institutions worldwide who have adopted predictive and generative AI-powered AML have seen transformational results in productivity, accuracy, and speed. However, Asian financial institutions lag their counterparts elsewhere in embracing these critical technologies.

“The rapid growth and varying levels of regulation and market maturity in APAC financial services present a unique challenge and an opportunity for organizations. Keeping pace with compliance demands a strategic embrace of AI with full board-level buy-in to drive meaningful change,” said O’Reilly.

Echoing his sentiments is Regulation Asia Co-Founder and Head of Research Bradley Maclean. Maclean explained that Asian financial institutions recognize the potential of AI for fighting financial crime, but the research shows a significant gap between ambition and action.

“The cost of inaction is rising rapidly. Financial institutions that delay AI adoption risk not only financial losses but also reputational damage and increased regulatory scrutiny,” said Maclean.

Meanwhile, Craig Robertson, financial crime subject matter expert, APAC, Financial Services at SymphonyAI also commented that AI is delivering both efficiency and effectiveness to FIs in the region.

“Financial institutions are using AI to detect new crime more effectively, reduce costly false positives, and control spiraling operational expenses. This proactive approach allows us to prevent crime instead of just reacting to it. The good news is, effective AI implementation can be incremental, delivering immediate value while paving the way for profound long-term transformation,” said Robertson.