SIU Today Fall 2017 - 19

Automated fraud detection:
the way forward
Fighting fraud is a manual operation within many organizations.
Thus, fighting fraud can be a time consuming and error prone
process. Organizations with an automated solution are usually
more effective at fraud investigation by directly recognizing claims
that need further attention or require active follow-up. Chances
of detecting fraud and limiting false positives to a minimum are
higher when detection methods are automated.

Challenging fraud is not a priority for IT
The majority of insurers indicate that they have a project planned
to better fight fraud. However, looking at the priorities from an
IT perspective, there are numerous projects that may well end
up higher on the list of priorities than fighting fraud.
Common themes for IT projects are digitization and replacement
of core insurance systems. So, it is up to the business to translate
their pain into IT priority. Moreover, why should insurers wait
with implementing a fraud solution? These projects can go hand
in hand as most solutions are easy to integrate in any core system
and should complement each other.

standards. Automating the process by scoring screenings via
a predefined model helps. The claims that score below a set
benchmark can immediately be paid out. Only claims that score
above the benchmark need to be looked at, either by senior
claim handlers or by a member of the Special Investigations Unit
(SIU). Automated tools will leave less room for fraud. Also, they
potentially enable a 100% STP of claims and policy applications.

Increase the effectivity of fraud investigations
By directly recognizing claims that need further attention or
require active follow-up, organizations with an automated solution
are more effective at fraud investigations. Because the correct
information is handed over to the fraud investigators, they can
spend their time on cases that actually require attention and
effectively build cases against the fraudsters. An essential aid for
an effective follow-up could also be the deployment of a Special
Investigations Unit. A great number of the insurers already have
such a unit in place or outsource the investigations. This does not
automatically mean that the others are not actively investigating
their potentially fraudulent cases. Insurers could also cover such
tasks in the current operations.

Outdated internal fraud systems
Unfortunately, a great number of insurance companies use
outdated internal systems or still rely on manual processes such
as knowledge workers or business rules for fraud detection in
Excel spreadsheets.
The follow-up and investigation of possible fraud cases is most
efficient if all related information is available in one system. It
would instantly be clear what cases require attention, whilst
the majority could probably be processed straight-through.
Investigating false positives will be reduced to a minimum as
objective risk assessment can be applied while following corporate
guidelines. Such an integrated experience makes working on and
sharing of cases less error prone as well.

Objectify screenings
Data of investigated claims is often based on investigations that
have not been carried out at random. There was always a reason,
feeling, or indication for starting an investigation on certain
cases. Such a feeling might be valid, but there is a chance that one
has searched for the justification of a preconception. Nowadays,
more and more techniques are developed to detect unwanted
biases. With an automated system, the influence of bias and/or
prejudice when investigating data is limited. This will improve
the quality of analyses. Thus, the managerial decisions that are
based on them will be better.

Leverage the advantages of STP
It is virtually impossible to screen all incoming claims manually,
let alone keep up with high and constant changing compliancy

Prevent rather than cure?
If an insurer is affected by a fraudster, the chances are high other
insurers are confronted with the same act by the same fraudster.
That's because in most occasions only the policy gets cancelled.

SIU TODAY | FALL 2017 | 19
SIU Today Fall 2017

Table of Contents for the Digital Edition of SIU Today Fall 2017
