Independent Banker - November 2024 - 59

consumer-permissioned
data or a combination-
each offers a way to make
the lending process more
streamlined for both the
bank and the customer.
" Machine learning models can
analyze check characteristics and
detect abnormalities that might signal
fraudulent activity. "
-Ravi Sandepudi, Effectiv
MACHINE LEARNING THAT
FIGHTS CHECK FRAUD
Year in and year out, banks
are forced to spend money
to keep fraudsters at bay,
whether it's payment fraud,
cybercrime or identity
theft. Since the pandemic,
check fraud has been on
the rise, with fraudsters
becoming very organized
and even breaking into
post offices to steal checks.
In 2023, the Financial
customers submit documents through a portal,
banks can use technology such as robotic
process automation (RPA), which is capable
of more complex automations involving
AI. RPA can extract data from submitted
documents, like bank statements, pay stubs
or loan applications, and send that data into
the bank's underwriting application or loan
origination system.
CONSUMER-PERMISSIONED DATA FOR A MORE
COMPLETE FINANCIAL PICTURE
An alternative to having customers submit
documentation is to use consumer-permissioned
data. Customers use account aggregation to
grant their lender secure access to data at
another institution, such as another bank.
Rather than relying on paper bank statements,
lenders can get a more complete financial
picture. Consumer-permissioned data can also
be used to grant lenders access to payroll data
for verifications of employment and income.
" Financial providers can't deliver the best
experiences for customers with only a fraction
of the picture, " says Wes Hummel, chief product
and technology officer at MX Technologies, Inc.,
a product that connects financial accounts. " By
enabling consumers to easily connect accounts
and share their financial data, both financial
providers and consumers can better understand
and do more with financial data. "
No matter how a bank opts to collect data-
whether it's through an online portal, RPA,
Crimes Enforcement Network issued an alert
about a surge in check fraud, particularly
mail-related check fraud. Suspicious activity
reports filed related to check fraud jumped
from 350,000 in 2021 to more than 680,000 in
2022-a trend that is expected to continue.
" Check fraud, though it may seem outdated,
remains a significant threat due to the high value
of transactions associated with checks, " says
Ravi Sandepudi, cofounder and CEO of Effectiv,
a risk management platform. " Machine learning
models can analyze check characteristics and
detect abnormalities
that might signal
fraudulent activity. "
Because of the speed
and volume at which
fraudsters operate,
community banks
should prioritize realtime
detection and a
solution that can do two
things, he says. One, it
should integrate with the
bank's existing systems,
and two, the technology
should be able to adapt
to new threats as they
of the U.S. banking population
prefers to do their banking on
a mobile banking app
Source: Forbes
emerge. Overall, check fraud technology reduces
the risks that come with manual intervention or
human error.
" If banks don't invest in effective
fraud-fighting solutions, they risk significant
independentbanker.org // 59
http://www.independentbanker.org
Independent Banker - November 2024

Table of Contents for the Digital Edition of Independent Banker - November 2024
