What is the Amazon Echo doing with my voice recordings?
I'm working on a project for school to determine what Amazon is doing with my voice recordings? I setup a test SSID with no security so I can see the packets and collected a capture. I can see the TCP packets but I'm having a hard time determining if the echo is turning the packets into audio or converting to text and sending them to Amazon for analysis? And, if they are audio, is there a way to replay the TCP packets and reproduce the conversation?
This whitepaper has a good description of the internals and how voice is processed.
Alexa Privacy and Data Handling Overview
https://d1.awsstatic.com/product-mark...
"The system is designed so that communication between the Echo device and the Alexa cloud is protected using TLS 1.2."
The data packets you captured are most likely TLS and there isn't an easy way to get the key to decrypt.
Thanks that helps. What's interesting is that the number of packets never really seems to change whether I have the "Help improve Alexa" options on or off. I expected a sharp reduction since it would no longer be using my voice recordings to "help Amazon," but at least from the packet side the amount of data being sent from my device remains unchanged.