Fraud in Five | How Analytics is Transforming the Anti-Money Laundering Marketplace

welcome to this edition of Fraud in Five. Anti-money laundering has been
one of SAS’s fastest growing markets over the
last several years. The market is demanding a better
balance between risk mitigation and operational
efficiency, underscoring the need for more robust
and explainable analytics. With me today to discuss how
analytics is transforming the AML marketplace
is David Stewart, our global lead in banking
for our fraud AML compliance and security solutions. David, welcome. DAVID STEWART: Thank you, Stu. STU BRADLEY: 2019 has been a
very active year for our AML customers. What is driving the
need for change? DAVID STEWART: Modernization
and innovation. Financial institutions
understand they need to modernize their AML
processes to be more efficient and to be more effective at
identifying complex risk. Digital transformation in the
financial services industry is driving a lot of this,
because they’re under pressure to reduce operational costs. And the days of throwing
bodies at the problem are over. So definitely an interest
in reducing false positives in an industry where there’s
about a 95% false positive rate on average. And customers are
really seriously worried about the
complex risks that they may be missing with traditional
rules-based systems. So in summary, our
institutions want help in modernizing with AI
and machine learning in mind, as well as moving to perhaps
moving some of their IT cost into a cloud or
managed service model. STU BRADLEY: Clearly,
regulators are impacting the need for change. How are they driving this
change in the market? DAVID STEWART: We continue
to see enforcement actions, but they’re more
reasonable and they’re more distributed globally. So the good news is,
there’s more parity in terms of enforcement actions. But I think regulatory agencies
are promoting innovation, as evidenced by the December
2018 joint statement on innovation by
the US regulators, which in summary says,
please innovate cautiously with AI machine learning. Run these programs
and pilot in parallel, while maintaining your
existing processes. So in general, regulators
are very encouraging around innovation. STU BRADLEY: So
what is SAS doing to help our
customers in managing through this
regulatory environment? DAVID STEWART: Well,
one of the areas that I think all
institutions struggle with, is knowing their customers. So in the case of some
of the New York branches we work with, their biggest risk
is correspondent banking risk, where their foreign parent is
introducing dollar clearing transactions where they
have very little information about the customer on
the other side of that. We’ve applied fuzzy
matching algorithms to create compliance
IDs so that we can create a profile on
the customer’s customer, at which point, we can
provide further analytics and anomaly detection. Of course, the other major
area of work that we’re hyper focused on, is helping
customers deploy machine learning into
their existing operations, as well as augmenting
to identify new risk. STU BRADLEY: The regulators
are encouraging innovation around AI and machine learning. SAS is investing over a billion
dollars the next three years into AI innovation. How are we helping our
customers adopt AI? DAVID STEWART: Gingerly. There’s a ton of hype in
the industry around AI. And we tend to focus on
providing commensurate risk coverage. We want to automate
certain processes. And obviously, with operational
efficiencies in mind. A few examples that come to
mind are analytical segmentation that help customers go from
red flag-based rules monitoring to behavioral monitoring. Another area is kind of
intelligent automation, where we automate the
escalation of alerts to cases. And then the third area
really, is for some of our more advanced clients. They’re deploying
SAS neural net models to replace traditional
incumbent rules. And in one mid-sized
US bank, we saw that they tripled their
SAR conversion rates and have cut their
work volumes in half. So that validates the approach. What we’re now doing is creating
a repeatable financial crimes analytics methodology
that can be rolled out to other clients that
are a little earlier in their journey. STU BRADLEY: Such
that we can have scalability and operationalizing
these capabilities. DAVID STEWART: Right. STU BRADLEY: At a massive scale. When you look forward out in the
next couple of years with AI, what is your prediction
think it’s a given that AI and machine learning
and robotic process automation will gain adoption in the
next three to five years. What we have to do as
a technology provider, is to make AI more explainable,
so the investigators know why they’re
looking at something. We have to automate some
of those mundane tasks, so that precious
human resources can be redeployed to investigate
those complex risks that create reputation risk
for the institution. And that’s what’s
exciting to me, is that hybrid
approach, a combination of human intelligence and
artificial intelligence. STU BRADLEY: Well,
David, thank you so much for the time and
information today, much appreciate it. DAVID STEWART: Thank you, Stu. STU BRADLEY: One
of my key takeaways from the current environment,
is that technology has outpaced a financial
institution’s ability to adopt. As such, as a vendor,
it is imperative that we make technology
easier to adopt, more transparent and
more explainable. Thank you very
much for watching.

Leave a Reply

Your email address will not be published. Required fields are marked *