Fintech2030 Q&A: The Role of Regulation and Technology in the Fight Against Money Laundering

Nasser Khodri is the Capital Markets Sell Side group president at FIS® and is recognized as a leading innovator and market commentator on the global evolution of capital markets technology.

Currently, only 1% to 5% of the UN-estimated $1.6 trillion to $4 trillion in laundered money is actually identified and seized. A lot of this is the result of there being so many false positives, with too many alerts for transactions that turn out not to be fraudulent. Firms simply can’t sift through them all.

Similarly, those laundering money are often five steps ahead of everyone else. They’re constantly looking for new ways to hide their activity. They’re using different tactics, mediums, and now even different instruments – think cryptocurrency – which makes it extremely difficult to detect when a transaction is illegitimate because oftentimes it looks legitimate.

Change also happens extremely fast in terms of customer processes. For example, it’s taking less time to open accounts and, therefore, to run identification checks on the people opening those accounts. While banks and other financial institutions are keen to shorten this time even further, it creates more opportunities for fraudulent activity to go under the radar.

What types of systems do companies have to catch money laundering?

The average large financial institution can have multiple different solutions across the enterprise, but often those systems don’t talk to each other, making it very difficult to catch money laundering.

For example, none of the data coming out of those systems can be used on an automated basis to enrich the data in the other solutions to enhance detection of anomalies. That can only be done via case management, where there’s a person pulling data out of all the different solutions and putting together the puzzle pieces. This means financial institutions can’t connect all of the dots to create a complete picture of suspicious activity when it comes to money laundering.

What’s more is that in financial services, you would be hard pressed to find a top firm that has not had an AI implementation project. There has been an overpromise and an underdelivery on a lot of those projects. This is because many providers are trying to develop an AI platform and strategy themselves, and that hasn’t worked.

As a response to AI projects failing, financial institutions have started throwing more personnel at AML.

What role is regulation playing in AML?

A promising regulatory step is U.S. Treasury Secretary Janet Yellen placing the creation of an AML registry on her agenda.

The cost of compliance is escalating and becoming a fundamental driver of a firm’s ability to turn a profit. This new AML regulation wants to create a central database. It touches on the problem of the cost of that compliance and the speed at which solutions providers can adapt and implement solutions to address that regulatory need.

Across the board, most firms will tell you that it took them a year or two to implement their current AML solutions. A new regulation means firms have to be able to use or adapt the existing solutions they have. If those existing systems are built on legacy technologies that can’t support change in regulation, they can’t comply within the needed time frame. Many firms are dealing with this.

Moreover, I think we’ll see a convergence of AML regulations. Bad actors are sprinkled throughout the entire market, and an institution looking for those bad actors on its own can identify only a very small number. But a centralized database where all of that data is then implemented and adopted provides a higher level view that can hopefully result in identifying more of the bad actors throughout the entire market and not just at a specific financial institution.

You said we’re only spotting 1% to 5% of the fraudulent behavior in the market. How can we capture the other 95%, and how can technology solve the AML problem?

The number one priority for firms has to be a broad adoption of new technology. The technology platform can’t be a secondary priority: It has to be something a firm does day in and day out. They also must add the financial services domain expertise to that.

Firms can no longer look at only KYC, transactions, and suspicious activity. Because of the rise of new technology and remote working brought on by the COVID-19 pandemic, everything is being pushed to new communications platforms. Even data from forums like Reddit can be used to give a more complete picture for fraud detection.

This means firms need to implement a holistic communications solution that examines all of their voice calls, text messaging, social media channels, internal emails, etc. alongside running their normal person and transaction checks.

I also think firms are going to have to become a little more aggressive on their approach to replacing or augmenting their current solution offerings. It is going to be a little bit expensive and disruptive. It’s going to be a balancing act between the customer experience and the backend operations. This is what we’ve achieved with our partnership with C3 AI®.

Can you give some details about this partnership with C3 AI and what you’re working on?

We have productized our financial services domain experience and C3 AI’s industry-leading – and most importantly, already proven – AI in a SaaS solution for AML to reduce false positives and increase suspicious activity identification.

We simply set our solution on top of any existing AML solution a financial services company may have. Sitting on top of all those different solutions, ours aggregates all of the data, connecting all of the different dots and provides analysis and insight into data that firms never had before. Our clients have already shown an 85% reduction in false positives and 200% increase in suspicious activity identifications.

Our solution also has the nimbleness to keep firms ahead of regulations. When a new regulation like an AML registry comes into effect, because our system aggregates all data, clients could, for example, simply connect an API to the central depository and automatically port that data over to the depository.

All third-party trademarks are the property of their respective owners and not affiliated with FIS. C3 AI is a trademark of C3 AI, Inc.