Slow Advancement of AI Poses Risks to Fincrime Defense, According to SAS Study

Although AI technologies are not a one-size-fits-all solution for every challenge related to financial crime, they have shown remarkable effectiveness in areas that deal with large amounts of data, such as automating alerts for transaction monitoring and reporting potential suspicious activities.

Professionals in anti-money laundering (AML) have pinpointed the reduction of false positives (38%) and enhanced speed and quality of investigations (34%) as two of the most significant advantages of implementing AI and machine learning.

SAS launched an AML technology study in collaboration with ACAMS this week, building on similar joint research released in 2021. This new report, which stemmed from a global survey of 850 ACAMS members, presented findings that indicate AI and machine learning (ML) adoption remains modest.

Just 18% of those surveyed have implemented AI/ML solutions in production, while another 18% are testing these technologies in pilot projects. What is even more striking is the fact that 40% have no plans to adopt AI/ML in the near future.

This report also found that Generative AI interest is robust but cautious. Almost half of those surveyed reported that they are either experimenting with GenAI (10%) or currently exploring its possibilities (35%), reflecting significant interest in this emerging technology. However, as mentioned, 55% of respondents do not have any plans to adopt GenAI.

Many of the conclusions from this study require change to happen with AI and its use in the workplace, but even if initiatives are done right away, it could take years to see the change unfold. The positive side of this is that these obstacles are manageable, and technology is assistive in enabling financial institutions to fulfill the regulatory requirements associated with transitioning AI from a pilot program to full-scale implementation.

In the next 5-10 years, there will be a shift in larger institutions creating innovative decision-making frameworks that enable them to detect and track activities in various sectors with significantly improved accuracy throughout the entire customer lifecycle. In order for this change to happen, though, AI needs to be trustworthy for banks to implement into their systems.

Without that trust and safeguards, unintended consequences can impact customers to a greater extent.

AI can feel unfamiliar and dangerous to those who haven’t used it before, but this study shows promises of it benefiting many banks and their databases.