NICE Systems this week launched its Contact Center Fraud Prevention solution, which tracks fraud patterns and screens all phone interactions against a watch list of known fraudsters.
Businesses that use the solution will be able to pinpoint fraudulent callers according to their unique voiceprints by using voice biometrics to automatically cross-reference each call with a suspicious watch list of known fraudsters.
The solution also uses NICE Interaction Analytics to identify fraud patterns and social engineering attempts based on speech analytics, emotion detection, talk patterns, and interactions. For example, shouting at the agent or trying to change an address or phone number could be part of a fraudulent behavior pattern, and specific keywords might raise a red flag. Telephony and other contextual data, such as IVR events, caller location, and caller ID matching, are also examined to determine potential fraud.
"The fight against fraud is essentially a high-paced technological arms race," says Ori Bach, director of solution management and interactions optimization at NICE. "As criminal organizations roll out new technologies and tactics to commit fraud, legitimate businesses are racing to deploy fraud prevention technologies to counter them. Many businesses have made major investments in the capturing and analysis of interactions to improve efficiency and the customer experience. This foundation can now be leveraged to understand and mitigate the risk of fraud during those interactions."
The NICE solution guides agents in real time to appropriately handle high-risk interactions and provides an end-to-end fraud management solution that prioritizes high-risk interactions for investigation before transactions are authorized. The offering leverages the NICE Actimize Risk Case Manager, a complete investigation and work flow tool used to open investigation tickets following identification of fraud attempts, including playback of suspicious interactions.
"Criminal organizations employ relatively small numbers of operatives that use similar social engineering scripts," Bach says. "Well-trained security professionals who routinely listen to fraudulent calls can often spot a bad call just by listening to the first few seconds of it. Some of our clients have developed such an intimate familiarity with the fraudsters and their gangs that they have given them nicknames. The challenge is that security professionals take a long time to train and are only able to listen to a small fraction of the calls. Our software essentially automates and speeds up this process by identifying who the speaker is using voice biometrics, what is he saying using speech analytics, and where the call is coming from using telephony analytics. This powerful combination is very effective in separating the fraudulent calls from the legitimate ones."