The Challenges of PEP and Sanctions Screening
The Challenges of PEP and Sanctions Screening
Politically Exposed Persons status (PEPs) and Sanctions Screening is an essential part of any company’s Anti-Money Laundering policy. But while it’s so widely practiced, the process frequently causes hurdles. So what does the screening involve and what makes it so complex?
Let’s start with the basic definitions:
Sanctions are regulatory or governmental orders that prohibit companies from carrying out financial transactions with a person or an entity that’s suspected of criminal activity.
A PEP is an individual who is or has been entrusted with prominent public functions either domestically or abroad, e.g.: a head of state or senior politician. Such persons are presumed to be at an elevated risk of corruption.
Businesses carrying out money operations (particularly financial institutions) have an obligation to check their customers against sanctions and PEP lists for potential matches.
Approaches to PEPs and sanctions screening vary depending on factors like the scale and nature of business. As a rule, however, PEP and sanctions screening can be split into 2 parts:
- Name screening
- Transaction screening
Name screening for PEP and sanctions is a critical component to any company’s AML policy. Exposure to PEPs and sanctioned persons, if discovered by authorities, may result in a fine that could prove ruinous to a business. Back in 2019, a top Swiss bank received a penalty of $5.1 billion for AML, KYC and sanctions violations. As it turns out, this sum exceeded the bank’s 2018 net profit of $4.9 billion. How does such an established financial institution end up with such trouble on their hands? Well…
The Possible Pitfalls
Name screening may seem like a straightforward task, but the reality is a bit more complicated. You see, the difficulty with name screenings lies in ensuring a true match. A true match means 100% certainty that your customer is indeed the person listed as a PEP or a sanctioned party. But how can one be sure of it when most database providers fail to include unique identifiers, such as a date of birth or an address?
Such are the limitations of outdated rule-based screening solutions. These legacy systems comb through client data with crude rules (i.e. basic matching criteria) with no regard or consideration for spelling, aliases, latest changes to databases and third-party references. As a result, the company performing the screening process either ends up with false positives that require additional investigation or runs the risk of not flagging a potential perpetrator.
The Way Forward
Fortunately, the solution to these pitfalls is already here. Artificial intelligence (AI) and machine learning are the two technologies at the helm the effort to improve screening rules. These two help fine tune detection and matching strategies to an astonishing degree.
For example, a subset of AI known as “fuzzy search” (aka “fuzzy matching”) enables companies to scan all name variations of the investigated subject, even those which may have been intentionally misspelt. It’s particularly useful for companies with international clientele with names in different languages and scripts.
Another invaluable AI-based process at the service of better AML is “adverse media screening”. It flags down any and all unfavorable information on the investigated person or entity that exists on the Web. This gives a deeper insight into their background and adds an extra layer of protection against associating with clients who have compromised themselves.
And finally, unlike the legacy systems that required manial updates and maintenance, the new AI and ML powered systems are fully automatic and use real-time data from a wider list of databases. DataSpike is one of the screening platforms successfully harnessing the power of AI and ML we’ve discussed here for better PEP and Sanctions screening. Our paired-back but powerful solution provides ongoing protection against AML breaches at a highly competitive price point.
Click here to learn more about our services or here to request a demo.