A murder investigation in Bentonville, Arkansas, has raised questions about privacy and smart devices inside one’s home that are always on. Amazon has objected to a subpoena from police seeking evidence from one of its Amazon Echo devices inside the home where the murder was believed to have been committed.
From the Washington Post:
While police have long seized computers, cellphones and other electronics to investigate crimes, this case has raised fresh questions about privacy issues regarding devices like the Amazon Echo or the Google Home, voice-activated personal command centers that are constantly “listening.” Namely, is there a difference in the reasonable expectation of privacy one should have when dealing with a device that is “always on” in one’s own home?
The Echo is equipped with seven microphones and responds to a “wake word,” most commonly “Alexa.” When it detects the wake word, it begins streaming audio to the cloud, including a fraction of a second of audio before the wake word, according to the Amazon website.
A recording and transcription of the audio is logged and stored in the Amazon Alexa app and must be manually deleted later. For instance, if you asked your Echo, “Alexa, what is the weather right now?” you could later go back to the app to find out exactly what time that question was asked.
Police did not specify what data they expected to find on Bates’s Echo — nor is it clear what the device could have captured that would have been relevant to the case. Only if someone happened to have triggered his device with its wake word would it have begun recording any audio. Even then, it seems unlikely that audio would be conclusive evidence of an alleged murder.