Voice biometrics system is used to identify and authenticate an individual using his or her unique voiceprints. Typically during the enrollment process, hundreds of voice characteristics including physical and behavioral features of the voice sample are combined to register a unique voiceprint by a software platform. The stored voiceprint is used to identify and authenticate the customer based on the calculated confidence rating when requesting access to perform transaction or application. Voice biometrics finds application for remote verification of clients over the phone. Some of the key end-users include financial services, telecommunication, immigration checks and so forth. This article aims to explore the reason behind the upsurge in the adoption of voice biometrics in financial services and telecom segments.
Financial Service and Telecom Sector
Financial services sector includes various segments such as personal banking, commercial banking, investment banking, insurance firms, credit unions, retirement solutions, and so forth.
Need for Voice Biometrics in Financial Service Sector
Voice biometrics has evolved as a trusted system to authenticate an individual for various reasons remotely. First of all, the market players need to solve the hassle faced by customers in remembering more passwords as well as, the effort required in trying to recover a forgotten password. Second, in an increasingly digitized world, people prefer non-branch banking such as the Internet and mobile banking, and much of the queries are addressed through call centers rather than branch banking. Last, financial services demand high security for customer assets, while the current method of authenticating a client via phone has proven to be unreliable. For instance, requesting access to the Internet and mobile banking has been strengthened beyond PIN passwords, through password token, SMS one-time passwords and so forth. However, authenticating a speaker over the phone through call center with “security questions” such as date-of-birth verification, number of accounts and cards holding and so forth is unreliable. From the customers’ perspective, the risk is that anyone with the account holder’s information can bypass the system, while from the company’s perspective, the process of verifying the customer is consuming a significant portion of time, which can otherwise improve the operational efficiency. With voice biometrics, while the customers speak with the friendly customer care executive, the unique voice print of the customer is analyzed and compared to the registered template within 25 seconds in the background. Voice biometrics allows the customers to be verified without the security questions, while the company saves close to 4/5th of the authentication time, improving the operational efficiency.
Some of the key advantages of voice biometrics include low cost of ownership, no requirement of special hardware, non-intrusive nature, remote, and secure identification capability of the technology.
Application Trends of Voice Biometrics in the Financial service sector
Insurance, retirement, investment, and banking firms such as Transamerica, and ANZ are some of the recent adopters of voice biometrics. Key reasons cited as a part of implementation include enhancing customer security as well as customer experience. Citigroup who was the first to roll out voice biometrics in APAC region finds that the clients can be verified within 15 seconds of the call duration in the background while they explain the reason for the call with the operator. Citi group has already enrolled close to 1million APAC customers in consumer banking division. On the other hand, noting the statistics that over 60% of the fraud initiates with a phone call, and at least $-10 billion loss was accounted for loss through call scams has driven the PSCU credit union firm to invest in fraud detection and risk management through a partnership with voice biometrics solution provider Pindrop. Another research study from Pindrop suggests that phone fraud has increased over hundred percent in 2016 on a year-over-year basis. Increase in technical sophistication among the fraudsters combined with the extensive amount of information available on the darknet has increased the attacks on banks, insurance providers and so forth. Similarly, a financial institution – BNP Paribas has integrated face, voice, and fingerprint recognition to its Wealth Management division after recognizing their customer wants virtual interaction and seamless connectivity, among other reasons. In a similar notion, several other credit unions and financial institution across the world are increasingly adopting voice biometrics. In summary, the primary driver for adoption of biometrics in financial services is attributed to enhancing customer security and experience, while improving the operational efficiency.
Telecom Industry Trends
In the past couple of years, telecom giants such as SK Telecom and Swiss Telecom have notably implemented voice biometrics. Even in Telecom segment, the primary motive behind upgrading their contact centers is to improve customer satisfaction and operational efficiency. Telecom verticals note that apart from reducing the handling time, they are also able to boost the NPS results. NPS values signify the customer retention and relationship. In summary, voice biometrics plays a significant role in helping telecom players take a competitive edge in the industry
Notable Technology Advancements and Challenges
In an attempt to improve the word error rate on par with human level, IBM researchers had reported a significant reduction to 5.5% from previous 6.9%. While the research suggests that there is a lot of work involved in further advancing the technology to reach the level, it is imperative that voice biometrics is viewed as one of the key requirement for Artificial Intelligence of the future. The financial services would want the voice biometrics to perform at a level, where chatbots will understand the conversation with the customer on an emotional level. Experts believe that to uplift voice biometrics into an attractive and leading solution; the technology should progress beyond the information exchange phase toward funds transfer and personal advice.
Challenges Faced by the Voice Biometrics Technology
The change in the voice of individuals over time at a differential rate is regarded one of the key challenges faced by the market players. Industry observes that variation in frequency and pitch of the voice over time makes the task difficult for identification. Aging, health, stress level, emotional state, and vocal effort are some of the contributing reasons for the change in voice. As a result of voice changes, the false detection or rejection would go up over time. To overcome the situation, market players use to strengthen the technology using spoofing detection and phone printing. Experts suggest alternative solution such as updating the model on a timely basis would lower these impacts; however, the system will be prone to security compromises.
In conclusion, voice biometrics looks very promising given their ability to improve customer satisfaction and operation efficiency. However, many technological advancements are required in this segment to make voice biometrics more valid and reliable in near–future.
