This article was produced in partnership with BAE Systems Applied Intelligence.
Bethan Moorcraft of Insurance Business sat down with Enda Shirley, head of compliance at BAE Systems Applied Intelligence, to discuss AML efforts in financial institutions.
Money laundering is often seen as an abstract white-collar crime that is far removed from the reality of most people’s lives. What many people fail to piece together is that the predicate offences linked to money laundering include some of the worst crimes like murder, sexual exploitation, human trafficking, kidnapping, terrorism, and organized crime.
According to the United Nations Office on Drugs and Crime (UNODC), between 2% and 5% of global GDP is laundered each year, amounting to as much as US$2.1 trillion. It is a severe threat to the global economy, and, as such, anti-money laundering (AML) is a critical area of focus for compliance officers across all financial institutions, including insurance organizations.
The ‘2021 Global State of Anti-Money Laundering Report’ from BAE Systems Applied Intelligence (BAE Systems) found that compliance regulations are creating a critical bottleneck for AML efforts in financial institutions.
To ensure a holistic view of the AML landscape, BAE Systems conducted a
The big problem with tackling AML, according to 92% of professionals questioned by BAE Systems, is the lack of collaboration between financial institutions, law enforcement and policy makers. The financial crime (FinCrime) feedback loop, or lack thereof, is hindering the progress and success of AML investigations.
“It’s really hard for industry professionals operating in a silo to actually evolve their FinCrime detection capabilities when there’s no feedback loop between financial institutions, law enforcement and policy makers,” said Enda Shirley (pictured), head of compliance at BAE Systems Applied Intelligence. “They’re simply seeing one side of the transactions, one side of the payments, and that is hindering AML detection. Until we solve that problem and improve collaboration, these bad detection rates are unlikely to improve.”
There is some progress being made both from a policy / regulatory perspective and from a technology perspective to improve AML within financial institutions. In June 2021, the Financial Crimes Enforcement Network (FinCEN) issued the first governmentwide priorities for AML and countering the financing of terrorism, which include: corruption, cybercrime, terrorist financing, fraud, organized crime, human trafficking, and proliferation financing. In doing so, the group recognized that more collaboration was needed among key stakeholders.
On the technology side, where BAE Systems is a market leader, there are “lots of really interesting things happening,” according to Shirley, particularly when it comes to federated learning, which is the decentralized form of machine learning. Federated learning is enabling individual insurance institutions to share what they’ve learned about AML in their siloes with their peers in the industry in a secure and anonymous fashion.
“Today, the industry works off basic static rules, and if we apply those rules to federated learning, it will determine if something is over a certain threshold and is therefore suspicious and should be investigated,” Shirley explained. “If insurers are able to collaborate and communicate with their peers in the industry in a sensitive way, they might be able to change that investigation threshold because they’re inputting data from a broader set. It gets really into depth and detail when you apply machine learning and analytics across the rules, and you develop clusters and complex definitions of scenarios, which you can then share with your peers in the industry in a secure manner – without any data being breached or impacted from a security perspective.”
BAE Systems provides financial crime, risk management and fraud detection and prevention for financial institutions, including insurers, via its
“NetReveal is like the central detection system for all the compliance and regulatory obligations required for customers across an entire insurance organization,” Shirley explained. “For example, if an insurer is onboarding a new policyholder, the immediate compliance obligation is to understand that prospective customer’s risk profile (using Know Your Customer checks) to determine whether or not to go ahead with the transaction or policy. The NetReveal platform facilitates those checks and builds out a holistic risk profile for each customer. Beyond that, the platform will then monitor each customer’s behavior over time and map that against criteria related to financial crime. If a customer’s activity triggers against those profiles, NetReveal will create an alert for an investigation to be generated.”
Read more: The evolving nature of fraud
Despite significant advancements in AML technology, the BAE Systems report found that 25% of compliance teams are still working with outdated technology, and a similar number (27%) claimed that investigators can’t keep up with alerts. Related to this is the fact that many respondents (29%) lack the resources needed to spot problems. With a focus on compliance as almost a tick-box exercise, many financial institutions are finding it hard to get the balance right between the roles that humans and technology should play in AML.
“There’s a complete mismatch and imbalance there right now,” Shirley told Insurance Business. “It’s phenomenal, the sheer amount of low hanging fruit that exists in terms of the mundane and administrative tasks that should be automated. It’s all about shifting the balance from administration to technical AML investigation. Compliance professionals want to detect financial crime – they have a moral obligation to do so – and yet, they spend all this time on compliance box-checking. Simply put, automation critically removes human error from AML investigations and can free up staff to focus on higher value tasks.
“In order for that to work, organizations have got to ensure that the data they’re feeding into this complex array of regulatory requirements is of high quality - and that comes back to the collaboration across the industry, law enforcement, and policy makers. Then, they must ensure that the actions they take are reflective of that high-quality data. By that I mean, they must understand their own AML policies, work with an automation-first mindset, and give as much space as possible to their passionate FinCrime investigators.”
