Voiceprint recognition is fast becoming as popular as facial recognition for security. Also known as voiceprint or voice biometrics, the technology recognises a customer’s unique vocal patterns to grant secure access.
It has some advantages over other biometric security processes like facial recognition, as a person doesn’t have to be physically present for voice recognition to work. Of course, this also opens it up to more vulnerabilities.
How voiceprint works
Voice security is set up by individuals who record themselves speaking, to allow a system to capture their unique voiceprint data. This is then matched to their voice in the future, so the customer will only have to speak a few words to gain access to their account.
It differs from the speech recognition technology that’s built into Amazon Echo, Google Home and smartphone assistants. This recognises what’s being said but not who is saying it. Voice recognition will identify who is speaking, but not what’s being said. Combining the two capabilities can provide a more streamlined customer experience when somebody rings a call centre.
Voice security in action
Indeed, some banks are experimenting with voiceprint technology to assist their customer support teams. Authentication is carried out quickly, allowing human staff to focus on the customer’s reason for calling instead of a longwinded security process.
Santander is trialling a system where voice recognition is replacing security numbers for phone banking. Customers can phone Santander’s telephone banking service and repeat a short phrase. This will then be analysed to ensure they are the right customer. The bank states that the software is sophisticated enough to differentiate between recordings and live speech.
Reza Attar-Zadeh, head of customer interactions at Santander, explained, “The human voice contains over 100 voice characteristics we can use to ensure that a caller is who they say they are, further protecting our customers’ accounts from fraudsters.”
Voiceprint technology offers a more secure solution compared to knowledge-based authentication. It will reduce call times (and abandonment) and is a lot harder to imitate compared to finding out someone’s place of birth and mother’s maiden name.
There are weaknesses, however, that dedicated hackers can exploit. Much like facial recognition, it is more effort to fool. One such attack involves using ultrasound waves that cannot be detected by the human ear but are picked up by smart speakers. Researchers managed to create ‘DolphinAttack’ to target popular voice assistants including Alexa, Siri and Cortana.
Another security glitch was uncovered by BBC Click reporter Dan Simmons, who was able to trick HSBC’s voice authentication using his identical twin.
Voiceprint data is also highly protected under GDPR. Meaning that organisations which use voice recognition must have robust data security and consent for its specific use. A leak of voiceprint information will have high ramifications. For consumer trust, reputation and the bottom-line.
Pros outweigh the cons
Still, the pros of voiceprint security outweigh the risks. It is sensitive biometric data and should be treated with the utmost care – with a particular focus on protecting and storing it. But there are many benefits for consumers and organisations when voiceprint is implemented well. Organisations that invest in voice recognition now, will improve their security, customer experience, call times and bottom-line. Now that’s something worth shouting about.
All content produced and published by Paratus People. All rights reserved. ©