Authors :
Deepali Handgar; Sumon Singh; Mehdi Rezaei; Dr. Jyotshna Dongardive
Volume/Issue :
Volume 8 - 2023, Issue 7 - July
Google Scholar :
https://bit.ly/3TmGbDi
Scribd :
https://tinyurl.com/ym4ckb6z
DOI :
https://doi.org/10.5281/zenodo.8211389
Abstract :
In 2020, 53.6 million Amazon Echo Speakers
(enabled with Alexa) were sold. This number rose to 65
million in 2021 [1]. Customers can engage with cutting-
edge technology in a more natural way by directly
dictating instructions to Alexa thanks to its user-friendly,
personalised vocal experience. The users' vocal
commands are translated into commands by the
vendor's cloud services after being transmitted over the
internet to the vendor's initial voice assistant devices or
companion applications for smartphones and tablets. As
this vast amount of data, primarily the user's personal
information, moves throughout the voice assistant
ecosystem, there are many locations where it is either
temporarily or permanently kept. As a result, it is simple
for a cyber-attacker to alter this data, raising serious
privacy concerns. This, study examines various assaults
on the Amazon Alexa ecosystem, including assaults on
speech recognition and processing on the cloud backend
and frontend audio capturing. We also discuss potential
attack surface reduction techniques to make Alexa and
other voice assistants more private and secure.
Keywords :
Privacy, Security, Alexa, Attacks.
In 2020, 53.6 million Amazon Echo Speakers
(enabled with Alexa) were sold. This number rose to 65
million in 2021 [1]. Customers can engage with cutting-
edge technology in a more natural way by directly
dictating instructions to Alexa thanks to its user-friendly,
personalised vocal experience. The users' vocal
commands are translated into commands by the
vendor's cloud services after being transmitted over the
internet to the vendor's initial voice assistant devices or
companion applications for smartphones and tablets. As
this vast amount of data, primarily the user's personal
information, moves throughout the voice assistant
ecosystem, there are many locations where it is either
temporarily or permanently kept. As a result, it is simple
for a cyber-attacker to alter this data, raising serious
privacy concerns. This, study examines various assaults
on the Amazon Alexa ecosystem, including assaults on
speech recognition and processing on the cloud backend
and frontend audio capturing. We also discuss potential
attack surface reduction techniques to make Alexa and
other voice assistants more private and secure.
Keywords :
Privacy, Security, Alexa, Attacks.
