Vendors
日本語

Time to Modernize Sanctions Screening

Create a vendor selection project & run comparison reports
Click to express your interest in this report
Indication of coverage against your requirements
A subscription is required to activate this feature. Contact us for more info.
Celent have reviewed this profile and believe it to be accurate.
3 March 2021
Neil Katkov

Screening technology is essential not just to performing sanctions checks on wires and payments transactions. Screening also underpins the entire AML value chain, from KYC/CDD and enhanced due dlligence to investigation and forensics. These broader processes around identification and risk assessment of customers have relied on highly manual investigation of screening hits, like painstakingly sifting through search engine results.

The inefficiency of these traditional processes is leading financial institutions to rethink how risk assessment is performed on entities. To do this, firms are turning to next-generation technologies including AI, RPA, and natural language processing to help automate the risk assessment process.

Ironically, though, the original use case of sanctions screening for payments is still mostly tied to the time-tested approach of name matching based on structured data analysis.

Sanctions screening has extremely high false positives rates, and all these false positives must be cleared. This has led to a relentless expansion of analyst teams, until they now number in the thousands at the largest banks. This is a huge operational burden and FIs are desperate for ways to re-introduce efficiency into the equation and bring these costs down.

Moreover, many fines have involved sanctions screening, including the largest-ever AML fine of $9.8 billion. This is serious stuff. And regulators are asking for more, like identification of beneficial owners.

There are two main challenges with traditional screening: inability to provide a contextual analysis of entities; and reliance on manual processes.

Traditional screening uses name matching against watchlists and at most keyword search of profile and adverse media datasets. But matching engines can’t understand the context of an entity’s appearance in adverse media; assessing the risk of these alerts remains a highly manual job left up to human analysts.

Even the alert management interfaces for sanctions screening are often still bare bones and not taking advantage of the improvements we see in case management modules aimed at investigation. Clearly, there’s a lot of opportunity to modernize the whole sanctions screening cycle.

A big driver for innovation in name screening is the maturation of digital financial services. Banks need to reduce friction and ensure a smooth customer experience. This calls for resolving as many exceptions as possible automatically, in real time. We’re seeing a lot of interest by banks in using AI, RPA and NLP to modernize KYC and CDD processes; and some extraordinary gains in efficiency using these technologies. It only makes sense to apply these next-generation techniques to the meat-and-potatoes realm of sanctions screening as well.

For more on how next-gen tech can help shake up sanctions screening, check out my new report, Transforming Sanctions Screening: Improving Performance with Advanced Technology.

Comments

Insight details

Insight Format
Blogs
Geographic Focus
Asia-Pacific, EMEA, LATAM, North America