Never-ending monitoring alerts. Flipping from system to system to find customer information, transaction details, and negative news. Documenting searches and investigation reports. SAR filings. Model validation. Annual risk assessments. Hiring new analysts to replace the ones that jumped to another bank for more money. Audit one month, regulatory examination the next. Constantly shifting priorities.
This sums up AML compliance for the past decade. No wonder every AML leader is looking to make programs more efficient. But is AML efficiency even possible? If it is, how?
What Is Inefficient About AML?
There are two primary causes of AML inefficiency:
- False Positives: More than 80% of transaction monitoring alerts are false positives. Same goes for sanctions, watch list, and negative news results. Yet all these false positives must be reviewed and resolved.
- Documenting case files with proof that every monitoring, sanction, and negative news alert was reviewed takes more time than the review of the alert in the first place. Endless screen shots, copying, pasting, creating PDFs, all add considerable time and is not improving the way banks detect and report suspicious activity.
Compiling all this information takes time. When this time is added up over the course of a day, a week, a month, and a year more than 60% of the work of alert analysis, case investigation and EDD reviews is actually copying, pasting, and “papering” files.
All jobs have components that are repetitive, necessary, and boring. But, when those components account for most of the work time, something is wrong!
We cannot however simply stop all the documenting. Managers need to review work, audit must audit and regulators must examine. The case file is the evidence each needs to assure the work and the decisions reached were reasonable.
Artificial Intelligence Solutions Are Mostly Just Hype
A day doesn’t go by without an article about the promise of Artificial Intelligence or “AI” and how it will transform transaction monitoring. It will be years before AI has a widespread impact on improving AML (more on that here).
Machines will not replace the need for qualified humans to make decisions about what is suspicious. Machines are a long way from replacing human brain function.
However, what machines can replace are monotonous manual tasks.
How to Fix Reliance on Manual Processes
Think for a moment what transpires when an EDD analyst searches for negative news:
- She must devise a search approach using key words she chooses herself or continually copy and paste a standard search string.
- She then scrolls through some number of results – how many is anyone’s guess.
- She opens some of the links.
- She may re-do the search if the initial results she sees don’t appear useful.
- She will click on more links.
- She scans numerous articles.
- She may decide something is relevant, but spends more time deciding what isn’t relevant.
- She then either takes a screen shot of the article or she copies and pastes the text into a Word file and then converts it to a PDF.
- And, on and on and on she goes.
Ponder for a moment, the inefficiency and risk of all this. Think of the time spent, over and over on this same process, and it is just one part of AML analysis.
This is where machines can help. Applications that automatically capture the images on an investigator’s screen, enable one-click report creation, and standardize alert reviews all accelerate work.
This is all possible, in fact it is happening today and this is where AML officers can begin to bring better efficiency now as we wait for the promise of AI to arrive at some unknown point in the future.