This post has already been read 801 times!
The future of Anti-Money Laundering
AML teams are human resource intensive with heavy manual tasking in the process. As more regtech comes online these teams are set to be streamlined to allow the ‘tech’ to do most of the heavy lifting.
The risk will always be leaning too heavily on early adopted IT and taking the element of human experience out of the equation.
FI’s are taking significant steps to identify suspicious behaviour within their IT functions and this has increased the volme of investigators to manage huge increases in alerts.
So many in fact, collectively SARS are overwhelming the regulators globally. These increases are driven by huge fines and regulatory pressures in a virtuous cycle.
AML is dominated by management of alerts and the sheer volume of data sifting. This is missing some of the largest frauds. Too much emphasise is being put on the data when more should be put on the actual investigation behind the data, with suspicious customers the focus not the transactions per se. A new intelligence software tool would solve this dilemma.
How to Investigate Money Laundering
Money Laundering Red Flags
Cyber Money Laundering 101
60 Best Money Laundering Research Papers, Books and web links.
What are the best Money Laundering Schemes?
CUM_EX DIVIDEND FRAUD EXPLAINED
What is a Politically Exposed Person?
Preventing Money Laundering 101
Anti Money Laundering Warning Signs
Red Flag Warnings for AML pt2
New investments in IT to manage legacy data issues will ease the pressure but the focus should then turn to investigating customers who are high risk with the data supporting that and not the other way around.
The large frauds uncovered by journalists external to FI’s with little in terms of powers to secure their data, have done so while still battling the data volume.
The paradise papers, Russian Luaunderomat and Panama Papers being examples of that. This is because the press follow the suspects not just the data, using the data to identify targets and then investigating the target. Without this there is no story.
This is the optimum way to uncover money laundering because it identifies predicate offences something an investigator will not find from the data alone.
It is because of this the advance towards an IT nirvana, where the software will identify suspicion, will not in fact stop money laundering. It will not identify the actual criminality, just suspicions of it. To complete an AML inquiry human investigators will always be required.
Automation of KYC
The constant rub for customers during the onboarding process relates to the time and questioning required to ‘vet’ customers appropriately.
Automated tools are coming online to streamline this, reviewing imagery of documents, linking to social media, IP addresses and digital identifications as some examples.
Yet all of this is easily sourced illegally or can be altered with little effort. FI’s that rely solely on software to perform KYC checks will become the subject of much debate post the next ML case showing the systems haven’t worked.
Today, know your customer (KYC) reviews are a time-consuming and manual process. Over the next few years, financial institutions will implement automatic tools to streamline the KYC process and gain a better understanding of their clients.
Financial institutions can utilize the technology outlined below to enhance their KYC process:
- Client is prompted to take a picture of a government-issued identification on their mobile device
- Financial institutions can then automatically authenticate that identification
- Client is prompted to take a selfie on their mobile device
- Financial institutions can then automatically validate that the selfie matches the picture on the government-issued identification initially provided
- Client connects social media account(s)
- Financial institutions will be able to understand clients’ interests, sentiments, personalities, social connections and interactions, and life events
When a client takes a picture from their mobile device, the financial institution will also receive the client’s location, device identification number and type of device operating system. With this new data incorporated in a machine-learning model, fraud can be prevented and reduced.
3. Machine Learning
The majority of financial institutions utilize fixed rules in their transaction monitoring systems to identify suspicious activity.
Only 2 percent of system-generated alerts result in a suspicious activity report filing with the applicable regulatory authority.
Utilizing machine learning will enable financial institutions to more effectively and efficiently identify suspicious activity. With machine learning, the computer learns as it is exposed to new data.
The computer can then identify suspicious activity that it has not been specifically programmed to identify. This is helpful in detecting anomalies that a traditional monitoring system would not be able to identify.
In addition, machine learning allows computers to be trained to risk rank alerts, allowing financial institutions to more effectively manage their compliance program.
Within the next five years, cutting-edge AML compliance departments will dramatically change how they investigate suspicious activity, conduct KYC and operate their automatic monitoring programs.
The majority of investigations will be automated, KYC will be conducted in a seamless manner and machine learning will more effectively identify suspicious activity.
The need for humans will move from conducting simple investigations/KYC reviews to operating as anti-financial crime specialists and providing guidance to their technology colleagues.
Contact us about our Intelligence AI solution.
|[contact-form-7 id=”61″ title=”Contact form 1″]|