News broke earlier in January that Facebook has acquired Wit.ai, an 18 month old startup that specializes in voice recognition technology. At first, this might seem like a strange move but upon closer inspection, the rationale is clear.
Millions of users are turning to mobile as their preferred platform, where typing long messages and interacting with friends is far more challenging than on a PC keyboard.
It’s clear that companies like Facebook face a challenge to make mobile interaction easier and more engaging.
Using Wit.ai’s expertise, Facebook can build a mobile-first platform with a voice activated interface and text-to-speech messaging some obvious steps.
The Facebook acquisition highlights the excitement and potential behind voice recognition technology. We are potentially witnessing a fundamental shift in the way we interact with our technology forever.
As we start integrating voice activated functionality into new smart devices and services we use on a daily basis, my primary concern isn’t one of convenience but of security.
As I wrote in this blog in September 2014, there is much work to be done in securing our digital devices from voice commands.
Most voice recognition technologies scan commands for meaning and then execute them. I believe there is a need for an additional step, one of authentication.
Does the person issuing the command have the authority to do so? When I ask the device to execute a command, does it validate that it is really me and not someone else?
As I demonstrate in the below video, it is quite simple to have a device act upon a voice command issued by a synthetic voice or by a 3rd party that has an access to the device – even remotely:
As Facebook and other leading companies add more voice activation technologies to their roadmap, it’s important to realize that we are also increasing the number of services and devices that are potentially vulnerable to voice attacks. So considering this, , let’s build it with safety in mind.
January 13, 2015