A London-based company has developed a new method for online identification and verification, focusing on a user’s face and voice instead of simple text passwords.
When asked to verify their identity, Facebanx makes a request to the user’s webcam or mobile device and a live stream is engaged. The software takes a still image of the customer, which is then added to a database of faces. The customer then reads the numbers 1-10 out loud. This voice detail is then appended to the customer’s face data.
Facebanx's CEO Matthew Silverstone said: “By simply pointing a device’s camera and taking a selfie and then repeating out loud four randomly generated numbers, customers will be able to log into their account within seconds. This technology is incredibly accurate as it combines two different biometrics into one seamless experience.
“Our product is proving incredibly popular to organisations in the banking, gaming, healthcare and retail sectors as it provides for the first time a level playing field against cybercrime. Now criminals will not be able to use malware to simply hack into customer accounts as they won't be able to reproduce a live streamed image speaking out loud randomly generated numbers. This is the first time that the customer has been handed power back to them to stop account takeovers from taking place.”
Select public trials are now underway at telecoms and banking companies, with Facebanx voice and face biometric details being added to customer accounts to provide heightened protection against criminal access to sensitive personal information.
Facebanx is a member of the FIDO Alliance (Fast Identity Online), an organisation of over 100 biometric and high-tech companies – including PayPal, Google and MasterCard – planning to improve online authentication with the first standards-based and open specification in order to diminish password dependency.