In terms of anti-money laundering (AML) and Know-Your-Customer (KYC) compliance, Artificial Intelligence (AI), in particular, has the ability to completely transform how banks perform these processes efficiently and effectively. AI is particularly valuable when performing repetitive tasks, saving valuable time, effort and resources that can be refocused on higher client-value tasks. Here are five key ways in which AI can help improve AML/KYC and client onboarding processes:

Accurate client risk profile and enhanced due diligence

The real power of AI lies in its ability to intelligently extract risk-relevant facts from a huge volume of data, but then to also synthesize and deduplicate that information so that it is both meaningful and concise. This allows unstructured data from different sources and formats to be classified automatically for the KYC profile. Once data collection has been automated, it becomes much easier to generate better risk insights, leading to more accurate risk calculations. This means that AI can automate the creation and updating of the client risk profile and match this against the classification process (i.e. high, medium and low risk) to ensure continued compliance throughout the client lifecycle. Furthermore, AI can make the process of identifying high-risk clients even easier for enhanced due diligence processes.

Ultimate beneficial ownership

AI’s ability to ‘read’ vast amounts of data (including unstructured text) and derive meaning from this can help in producing comprehensive, accurate and auditable risk profiles on companies and individuals in a matter of minutes. This can add huge benefit to compliance teams who are tasked with weaving through complex webs of data on shareholders, beneficial owners, directors, associates and will improve their ability to draw accurate conclusions for a risk-based approach to compliance.

This will gain even more significance over the coming years given the enhanced global focus on the identification and ability to perform customer due diligence on ultimate beneficial owners in the wake of the Panama Papers scandal and the establishment of national registers to improve transparency in this area.

AML Screening and investigation

The current state of KYC and AML requires manual investigation, especially at the alert investigation phase, which is costly, time-consuming and prone to error. Today we take a blanket approach, where every alerted transaction requires human interaction. A recent Dow Jones-sponsored ACAMS survey reveals that the area of false positives is one of the most challenging for bank compliance teams (8).

Underpinning the alert generation process with AI can result in fewer false positives, for example, by deploying linguistic techniques to undertake watch list management (e.g. OFAC lists) that can vary languages and scripts (spellings) and can listen to news feeds to identify people not currently on industry watchlists.

While they are a significant part of the AML compliance process, alerts are not enough to support an effective and thorough investigation process. What is required is the linking of high quality data to the alert (via interpretation and link analysis) to produce an accurate, graphical representation of the legal entity structure. AI can help to leverage previously performed steps in the alert investigation process to formulate a recommended next steps approach.

Improved client onboarding and document management automation

The banking industry would dearly love to move away from being so document-centric, however, this won’t materialize for a while yet. Instead, we need to switch our thinking to how we can add and extract value from documentation. In today’s banking world, much of the client documentation involves scanned documents, which means a best guess approach needs to be deployed to categorize them appropriately.

When applied to workflow automation, AI has the ability to transform the generation of documents, reports, audit trails and alerts/notifications. AI’s Natural Language Processing (NLP), which allows it to ‘read’ vast amounts of information in any language, can enhance the KYC process for new client onboarding applications through intelligent document scanning and its ability to sift through a vast array of external data sources. This can significantly improve the overall client onboarding experience.

Managing regulatory change and compliance

AI’s ability to detect patterns in a vast amount of text (even unstructured text) enables it to form an understanding of the ever-changing regulatory environment. The panacea is auto-monitoring and interpretation. In other words, to automate the monitoring of regulations, transform them into a structured language and creates an ontology that allows the codification of rules with full traceability. In doing so, it has ability to track the changes in regulations around the world, identify gaps in customer information stored by the bank and provide KYC alerts to perform a regulatory outreach to clients to collect the outstanding information.

Furthermore, NLP can analyze and classify documents, extracting useful information such as client identities, products and processes that can be impacted by regulatory change, thereby keeping the bank and the client up-to-date with regulatory changes.

Conclusion

In a world of increasingly digitally sophisticated customer experiences, banks are now no longer competing against other banks – they are competing against every service experience in the world. Therefore, the stakes are much higher than they have ever been. This will fuel rising demand from banks for new, disruptive and pivotal technologies that will help deliver the Holy Grail of banking – happy, compliant, revenue-generating customers.