Amazon’s Advanced Voice ID

Alexa might be learning exactly who you are

It looks like Amazon is at it again with yet another new disruptive project that could change the way that Echo (plus all of the applications associated with it) interacts with users. Instead of responding to all commands, Alexa will be able to distinguish between individual voices with Voice ID. Amazon has reportedly been working on the feature since 2015, although the company has not yet confirmed or denied the project. Considering the emerging trend of audio-transmitted data, it’s difficult to believe that the aggressive expansionists aren’t working on an advanced voice-enabled service, especially when the primary form of communication used by Alexa is speech. So, who else is working on voice recognition technology, and why is it important?

Voice recognition technology. . . what’s the point?
Amazon are one of many companies reportedly working on voice recognition technology. Barclays currently uses customer’s voices as verification for over-the-phone transactions, and Google Home can work out whether a Pixel user has issued a command. The aim of advanced voice recognition is to achieve seamless identification, without the need to answer countless security questions or physically type personal information. This has clear advantages in a society which runs on passwords. Unless you use the same passwords and information, it’s impossible to keep up – and that really isn’t a good idea. As well as improving general efficiency, advanced voice recognition also acts as a foolproofing measure. In a recent (and hilarious) blip, a six year old girl asked Amazon Echo if it could play dollhouse with her. The device promptly ordered a giant, expensive dollhouse, along with four pounds of sugar cookies. After the story aired on a local news station, Alexa was triggered to make even more orders. As amusing as this is, it’s a reminder of how easy it is for unauthorised users to take advantage of the device, knowingly or not. Whilst it’s true that owners can install a four digit security code, this is not a foolproof solution. By using someone’s voice as ID, you essentially give them a vocal fingerprint.

How disruptive is advanced voice recognition?
Knowing who is making a certain query is obviously very useful in avoiding situations like the dollhouse fiasco. By identifying users, Alexa and its associated skills can be certain that they are completing a genuine order for the right person. Alexa now supports 10,000 ‘skills’, or, in other words, compatible services. Most of these applications involve shopping and purchases, and it’s in retail where voice recognition has considerable disruptive potential. For example, if an app can recognise someone’s voice, they can send super-targeted ad campaigns to that person alone. Team this with user preferences, and Alexa could work out what you want to buy even before you do. In finance, advanced voice recognition has already been adopted – but development in the reliability of the technology will encourage other banks and financial companies to swap out lengthy security procedure for vocal verification. Talking of sonic barcodes, London based startup Chirp are using data encoded audio to send information. Integrate this with accurate voice ID, and you could transmit reams of content via a single conversation. In the domestic sphere, Google Home already has some level of voice recognition technology that recognises when a Pixel user says ‘OK Google’. However, if Amazon’s Alexa can distinguish between people and commands, then it will undoubtedly disrupt and consequently overtake its Silicon Valley rivals. The changes caused by advanced voice recognition will impact security in all industries and applications. Adding another layer to security settings is no bad thing, and could act as a safeguard against hackers and cybercriminals.

Advanced voice recognition has clear benefits and will bring positive changes to transactions within retail, connected homes and finance. It will also help to improve security settings, especially in innovative technologies like smart homes and autonomous cars. The successful adoption of these devices relies on their ability to connect and respond to human commands – specifically to those of their actual owners. Individual voice recognition isn’t exactly a foolproof solution, though. It could be possible to replicate someone’s voice, or force them to give certain commands to gain access to property or information. Although these situations are extreme, they present a convincing argument for the continued use of ID codes and passwords alongside new security settings. When it comes to personal information in a world run by data, you can never be too careful.

Could your business use voice recognition security? Which other industries could take advantage of voice recognition? Is it a reliable form of identification? Share your thoughts and opinions.