Security concerns with voice activated devices

[ERDrPC]

Check out this article from wired

https://www.wired.com/story/lasers-hack-amazon-echo-google-home/

Basically using a simple laser, voice assistants can be triggered from distance to follow commands. 

Supposedly Google and Amazon are aware and working with the researchers who discovered this major flaw. 

[Cyknight]

11 hours ago, ERDrPC said:

Supposedly Google and Amazon are aware and working with the researchers who discovered this major flaw. 

No, they have stated that they are looking at the research to see if it holds merit.

This is quite farfetched. Just because you can create a virtual 'sound' by pointing a laser at a microphone does not mean you can trigger commands. Nor does it mean that a lab created environment is in any way usable in real life. Not to mention, the person wanting to do this would need to be able to see the device from a window.

And modulating a laser frequency isn't exactly something you can do with a laser pointer (the whole use of that device as an example is non-sense as such - it jsut points out you don't need a crazy powerful laser, but you still need to be able to modulate it).

 

It's neat science experiment, but this is hardly a concern for safety. Especially as you shouldn't be UNLOCKING (or disarming) anything using voice commands to begin with. You'd already have to be bypassing several security measures yourself to do this as it stands.

 

It also comes down to this: even if this is an actual feasible way of breaking into your house...the effort involved is so much more than just kicking in your door. And the scope-out required is more work than learning where your nearest security keypad is, so no, the whole idea of doing it 'silent;y' without raising an alarm doesn't work either.