The integration of digital assistants like Siri, Cortana or the new Google Assistant, have come to facilitate the life of many mobile users in their search for a dynamic, agile and friendly in the management of their devices, however, is that they now also represent a latent threat to the security and integrity of our information.
A recent research carried out by engineers from the University of Georgetown just check the possibility of developing a method of intrusion into any device with a personal assistant as above them indicated, through the inclusion of a series of simple voice commands with your audio modified at a frequency that can be recognized by any smartphone. As shown in the video:
View on YouTube
The presented proof of concept was an Android smartphone with the latest version of OK Google, ran a video with the modified audio file, under a modulation where the human listener could not understand the voice command, but the phone Yes, achieving control, indirectly various functions of the device, without the user noticing.
The document, scheduled to be presented in its entirety during the next USENIX Security Symposium in Austin, Texas, directly poses a scenario where this audio could be inserted even into YouTubevideos, activated via another malware already installed for different users, from revealing its location by GPS to lock the terminal damage.
No one is safe, not exempt of paranoia.
