Printer Friendly


Fraud is on the rise, and contact centers are a target. It's time to fight back. Be proactive and outsmart fraudsters with technology.

The latest advances in machine learning techniques and analytics combined with biometrics capabilities takes contact center protection to a whole new level. Contact centers now have the ability to expose previously unsuspected fraudulent activities. That empowers them to not only prevent known fraudsters from attacking, but more impressively, expose new fraudsters that were operating "under the radar" for a long time. In the ongoing response to sophisticated fraud attacks, the creation of high-quality fraudster watchlists from historical recordings--and automatically populating and updating them--is a revolution in the fight against fraud.

The Contact Center is Under Attack

The contact center is often referred to as the gateway to customer centricity. It is the point at which organizations achieve meaningful, lucrative connections with customers.

Fraudsters have a different perspective. They view the contact center as an opportunity for exploitation.

It is the weakest link they can leverage for their nefarious intentions.

Live calls from fraudsters are often orchestrated to appear as seemingly innocent inquiries. Instead, they are part of a wider campaign to systematically learn about an organization's security processes.

They're used to augment data the fraudster may already have obtained through other methods such as large-scale data breaches. Indeed, these breaches continue to fan the flames of fraud, as billions of PII (Personal Identifiable Information) credentials were reported as stolen in 2018 alone.

Often, syndicates use call-miners who are tasked with validating the information sourced through these various means. These data miners (also referred to as data mules) are extremely skilled at socially engineering live agents and often tend to be more pleasant to deal with than the generally irate, legitimate customer, thereby leading to even greater success through agents who are eager and incentivized to please callers.

Even fraud that is perpetrated outside of the voice channel--such as on a website or app--often relies on data that is validated at some point through a contact center. In fact, 60% of all of account takeovers involve the contact center, according to Forrester.

Fraudsters know they can finesse their way past voice channels and their human gatekeepers to exploit the data these organizations are supposedly collecting in the name of customer centricity, substantially hurting organizations, employees, and customers in the process.

Conventional fraud prevention solutions are not good enough. Its time to step up the game and adopt new technologies to stay one step ahead of fraudsters.

Organizations Must Be Proactive and Take Action. Now!

Organizations now realize sophisticated fraud attempts are a matter of WHEN rather than IF. No one is safe.

But the solution is not simply investing more heavily in security. Many security solutions are inefficient and ineffective. With numerous innovations continually being introduced to improve security while striving to also improve operational efficiency, organizations might feel bewildered with options. Even though multi-modal, multilayered security architectures are the desired end-state design of most security professionals, the time, cost, and resources are often prohibitive to most organizations. They subject organizations to enormous costs, while actually rendering the contact center riper for attacks.

How can organizations choose the fraud prevention solution that best fits their strategy?

Utilize the Newest Disruptive Technology

Because the voice channel is the attack vector of choice for fraudsters, voice biometrics is the technology best suited to prevent fraud in the contact center.

By using voice biometrics technology, an organization can strengthen security without increasing customer effort or frustration, because the process is automated, organic, and seamless.

Voice biometrics also improves agent experience. By eliminating a redundant task, it helps agents focus on the work they tend to do better, which is providing the best service. And because the automated process is more effective, the agent experience gain does not come at the expense of security.

Combining voice biometrics with the latest technology advances takes the protection of the contact center to a whole new level.

Adding powerful interaction analytics tools like Natural Language Processing and Phonetic Speech Analysis to voice biometrics help organizations gain insights about the behavior of callers based on what they say in an interaction and how they say it. This helps detect fraudulent behavior and flag suspicious activities. In addition, a robust fraud prevention solution must have the ability to analyze complex call scenarios typically used by fraudsters, like calls "on behalf of" someone else or calls where the speakers changed.

However, identifying suspicious interactions en masse is just the first step in proactively exposing fraudsters. New technology advances in Machine Learning and Deep Neural Networks make it possible to process all the data from these interactions, as well as data provided by other sources (like agent screens, CRM, and more). Plus, they can automatically asses the risk of fraudulent behavior in a way that is a lot faster and much more advanced than manual human assessment. For example, if the nature of the transaction in the interaction has changed during the conversation with the agent to a higher risk transaction, it will automatically be flagged as suspicious and analyzed to asses the risk of fraudulent behavior. This behavior would probably not be detected if the solution relies on manual work, like the agent flagging the transaction change for further investigation.

Combining all these technologies is used for automatically creating and maintaining a watchlist of fraudster voices (also referred to as voiceprints) from the suspicious interactions. Using voice biometrics technology, fraudsters will be identified and blocked in real time when they call the contact center.

Expose Unknown Fraudsters

Flagged voiceprints are fueling a growing watchlist of confirmed and suspected fraudsters. However, one of the challenges is the pace at which these watchlists are built and maintained.

Another challenge is the extent of resources required to perform analysis on the voice channels which, as with all fraud management strategies, stretches the limited resources of investigation teams to perform quality and speedy investigations.

Historically, the voice channel was used to supplement fraud investigations by manually accessing a history of recorded calls, physically listening to those calls to interrogate the content of the conversation, and performing a manual, qualitative assessment of whether the speaker sounded suspicious or different to the voice of the known or confirmed customer. Not only is this process extremely subjective, but it's also very time consuming.

Moreover, an investigation would be initiated only after a fraudulent activity already occurred that raised the suspicion of the investigation team (for example, following a security breach, or a customer filing a complaint their account was jeopardized).

The latest developments in a combination of speech analytics, voice biometrics, and machine learning has made it possible to analyze large databases of historical call recordings to link voices to each other, thereby creating high-quality voice profiles for all historical callers.

As the majority of fraudulent behavior is performed by a relatively small set of perpetrators, the challenge is identifying these individuals among the myriad of data sources and voices, inputs, and ingenious methods being applied to disguise activities and identities. If history does indeed repeat itself, the same holds true when it comes to fraudsters calling a contact center, whether it is a "brute-force", repetitive attack on a particular account, or a systematic collection of information through "vishing" of multiple accounts.

By automatically analyzing these recordings in-bulk and supplementing them with additional behavioral, transactional, and authentication data, ground-breaking results have been delivered in identifying known perpetrators as well as exposing previously unsuspected fraudulent activity.

Automation of this process allows companies to maintain pace with the growth in contact center fraud, not only at the point of attack, but for multiple calls that lead to the final act. The ability to leverage historical recordings delivers extremely rapid return on effort, as it leapfrogs incremental methods that rely on an organic ingress of calls.

Time to Be Proactive

In the ongoing response to sophisticated fraud attacks, the creation of high-quality fraudster watchlists from historical recordings provides a highly effective defense.

Accelerating the quality and scale of fraudster voice-watchlists therefore delivers exponential benefits in efficiency and accuracy.

Combining it with powerful interaction analytics tools and deep neural networks allows organization to expose unknown fraudsters that already penetrated the defenses and stop them from causing future harm.

Removing these "bad actors" from the picture is a revolution in the fight against fraud.

Fraudsters are creative and relentless. They will not stop because they hit one firm line of defense in one particular scenario. To truly transform the fraud prevention process, organizations must adopt a proactive approach.

They must have an ability to identify and eliminate any threats before they materialize.

NICE Real Time Authentication (RTA) provides a Proactive Fraud Prevention approach as well as real time fraud detection using a watchlist. It allows contact centers to take proactive steps in protecting their consumers by stopping fraud in its tracks immediately upon deployment.


* Contact centers are under attack, and fraudsters target the weakest link: the agents.

* It's time to act: don't wait to be attacked. Be proactive.

* Utilize the latest technology to outsmart fraudsters:

Voice Biometrics, Interaction Analytics, and Machine Learning.

* Expose and stop unknown fraudsters that already penetrated defenses.
COPYRIGHT 2019 Information Today, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Sponsored Content
Publication:CRM Magazine
Date:Jul 1, 2019
Previous Article:AI Gives Sales Teams a Boost: Enablement and engagement technologies are taking the tedium out of sales processes.
Next Article:MaritzCX Energizes Reliant's Customer Experience: The company has experienced success by adopting a customer-centric DNA.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters