Preventing Fraud in the Call Center with Phone Printing and Voice Biometrics, a recent Forbes report states, “Most call centers operate with one primary goal: customer service. Unfortunately, this leaves them vulnerable to “bad actors” and fraudsters who perpetrate crimes by socially engineering call center agents who typically employ weak methods to authenticate callers.” Current authentication methods that frequently fail include knowledge-based questions, PINS and Caller ID.

Voice is a much stronger form of caller authentication than traditional security questions and account information, and according to Gartner, shortens calls by an average of 20 seconds.


Voice biometrics verifies callers’ identity through the unique patterns in the voice. The technology is experiencing rapid growth, especially in contact centers hit by fraud and identity theft. Banking, insurance, and healthcare providers are using voice authentication to prevent fraud and to speed up the caller verification process.

US call centers spend over $12 billion each year verifying the identity of their customers. Voice authentication translates into huge savings for these organizations and a better customer experience, making it a fast-growing, industry-wide, technology trend.

Correct Terminology

“Confusing” might be the most accurate way to describe what is sometimes written and said about voice biometrics, as it applies specifically to the contact center. Two or more terms that are not really interchangeable are being used to describe technology that acts as a “gatekeeper” for access to a secure system like telephone banking. Would you be more correct to call it “speech recognition” or “speaker recognition”? Actually, we would be more specific and correct if we referred to it as “speaker verification” or “speaker authentication,” as opposed to “speaker identification.”

This “beginner’s glossary” comes from Wikipedia:
speech recognition – recognizing what is being said
speaker recognition – recognizing who is speaking
voice recognition – used for both of the above

In addition, there is a difference between the act of authentication (commonly referred to as speaker verification or speaker authentication) and identification.

Speaker verification is usually employed as a gatekeeper in order to prevent fraudulent access to a secure system (e.g. telephone banking). Systems like these operate with the users’ knowledge, and they typically require user cooperation. Speaker identification systems can be implemented without the user’s knowledge to identify talkers in a discussion, alert automated systems of speaker changes, check to see if a user is already enrolled in a system, etc.

Speaker Recognition Systems

Each speaker recognition system requires two steps: Enrollment and Verification. During Enrollment, the speaker’s voice is recorded and a number of features are used to form a voice print, template, or model. During Verification, a speech sample is compared against the previously created voice print. Capture of the voice biometric is considered non-invasive. Voice prints are created from some 200 characteristics of a person’s speech, providing a unique form of identification similar to tracing the patterns of a person’s fingertip. It is an easy and almost invisible technique. It is cost-effective because the technology typically uses existing microphones and voice transmission technology, allowing recognition over long distances via ordinary telephones (wired or wireless).

In May 2013 it was announced that Barclays Wealth would use speaker recognition to verify the identity of telephone customers within 30 seconds of normal conversation. The system used was developed by voice specialists Nuance, the company behind Apple’s Siri technology. A verified voiceprint would be used to identify callers to the system. The private banking division of Barclays was the first financial services firm to deploy voice biometrics as the primary means to authenticate customers to their call centers. The system was rated at “9 out of 10” for speed, ease of use and security by 93 percent of customer users.

Knowledge-based security has become progressively more complex. When one level became ineffective, another level was added. PINS led to passwords and passwords then required security questions to verify them. These authentication methods became complicated and time-consuming for the customer, and are easily compromised. The four-digit PIN is the weakest gatekeeper. It’s often shared, and a cyber hacker can quickly compromise it. Passwords and security questions can be successfully answered with simple web searches of the account holder.
Voice biometrics eliminates much of the risk by using the customer’s unique voiceprint for authentication. It can be passive, allowing the user to say anything and their voice can be matched to a voiceprint. Or it can be active, requiring the caller to recite a passphrase. Either method is an easier and much more accurate way to authenticate caller identity.
A voiceprint is a hashed string of numbers and characters so a compromised voiceprint has no value to a hacker. Not only that, each time hacker speaks to an IVR system, call center or mobile app, they leave their own voiceprint that can be used to proactively keep them out of the system and even alert law enforcement.

How Voice Biometrics Is Being Used

Voice biometrics is increasingly used in businesses and governments around the world for security, the protection of data and for increased customer convenience.
According to the Verizon 2013 data Breach Investigations report, “76 percent of network intrusions are from the exploitation of weak or stolen credentials.” Banks are particularly interested in improving the security of corporate transactions and customer access to account information.
When Barclays Wealth began to use voice biometrics with their international banking customers, the software allowed them to cut verification time by up to 20 seconds and to direct their primary focus to the client’s needs, rather than on the authentication process. Because of its success with its international customers, Barclays has expanded the technology and, since mid-2012, has offered it to the rest of its 12 million retail customers.

The global market for voice biometrics is expected to reach $30 million by 2020, with the Asia-Pacific region growing fastest, at a CAGR of 21 percent over the next five years. Governments and corporations have already compiled at least 65 million voiceprints in databases for security, surveillance and commerce and industry analysts suggest this figure could more than double over the next two years. The Associated Press reported that the voice recognition industry could generate as much as $900 billion in revenues in 2015.

With the continued rise of identity theft and also the desire by companies to securely help customers access information as rapidly as possible, “voice biometrics” has emerged as a way to meet these challenges and secure customer data.

For more information on the Voice Biometrics Industry see the latest market research: Voice Biometrics Market.