Do you have a code language with your phone’s virtual assistant?
Well, if you haven’t done it so far then someone else just beat you to it.
And that someone is surely not a well-wisher.
Imagine the things they can have your virtual assistants do while you look for a fault as to why your phone is working up.
With the phones getting more advanced than ever and all the talk about the untapped potential of a virtual assistant, it seems that harm can be done way beyond your imagination.
They can take over anything electronic, your phone, your tabs, your watch, your gaming console and even your car.
Sounds dangerous, doesn’t it? It is.
How do they do it?
Simple biology.
The hearing capacity of a human being is from 20 Hz to 20 kHz.
Some creatures can listen to voices louder or less than this frequency.
Just like the silent whistle dog owners blow for their puppies, dogs can hear it but not the human blowing it due to the limitation of the human ear.
And so we have our microphones that can catch voices above 20 kHz and thus listen to sounds that we can’t.
This can be done easily using the method called “Dolphin Attack.”
What is Dolphin Attack?
It is a method developed by researchers at Zhejiang University in which they use ultrasonic sounds to communicate with the best known virtual assistants.
Here is how it works:
- The person (most probably a hacker) speaks a command.
- That command is then translated by Dolphin Attack into ultrasonic sound waves.
- The sound is then sent to the device
- And the devices respond.
How can the hackers send this command?
That’s the question, isn’t it?
This kind of exchange can only take place if the device is in a proximity of the hacker.
Now. Not to make you paranoid, but you never know what the strange guy sitting on the coffee table next to you is doing.
That’s not even the most dangerous one.
Imagine commands like these embedded in your TV commercials, or the pop up annoying windows that play on their own, etc.
Does it work on all the virtual assistants?
Researchers tested it one Google’s assistant, Siri, Cortana, and Alexa.
They all responded very well to attacks in English, Chinese, German, Spanish, and French.
How can we prevent this attack?
Well, the most immediate solution is to switch off the voice command.
This will leave the virtual assistants to be operated only manually.
In the long run, companies can work on their virtual assistants and program them so that they only capture sounds that can be heard by humans.
This is no way anyone will be able to mess with the phones right under the owner’s nose