From a technical viewpoint, it’s extremely likely that the local device, the echo or the Google Home or the dot, is like old security cameras used to be, in that it just keeps about five minutes worth of sound and keeps overwriting that as it goes, since all it’s listening for is the wake word. If it doesn’t hear the wake word, It doesn’t need more than that.
Both Amazon and Google have said specifically that nothing is sent to their servers unless the wake word is recognized.
So I’m pretty sure that the law enforcement officers who requested these Records are going to find that they don’t actually exist. That the only thing that is on the servers are the specific commands that followed the wake word, which my understanding is what Amazon did turn over after the appropriate warrant was received.
“Always listening” isn’t the same thing as “keeps the records of what it heard forever”
But we’ll see.
These headlines strike me as little more than sensationalist click-bait (either intentionally so, or due to technical ignorance on the parts of the authors…something depressingly common even among consumer tech-oriented publications these days). There is nothing in any of these pieces to to substantiate the breathless claims made in said headlines (at least the first two), or that this case amplifies anything other than the fact that the Bentonville PD doesn’t know how the Echo works and is grasping at straws.
And more fodder for the tin foil hat brigade.
Of course they are. They are fishing for evidence. Sometimes it pays off. Sometimes it doesn’t.
And consumers want to know the results as well.
Amazon declares the way it works, and if 100% accurate, nothing will come of the subpoenas but this is news worthy because certain consumers are interested to know if the weight of our judiciary can ferret out further privacy issues that are not currently known to the public. Is that also grasping at straws? Sure, a bit - but sometimes you get one!
We’ll see. Either way the results will be interesting and noteworthy in my book.
Yea, folks wanting to know if the hot cloud enabled mics they welcomed into the most intimate parts of their lives are a privacy burden beyond what they know or even a legal liability.
Totally Tin Foil to want to know these things… what a bunch a worry warts!
I think you’re putting words in his mouth…virtually speaking. He didn’t say that those with rational questions are paranoid. He said that there are those who are paranoid about this issue and that this story will feed that paranoia…and he’s demonstrably correct. I read several forums that, the moment the story broke, were chock-full of such people declaring, “See! I told you these things were spying on you!” because they not only don’t know how the Echo works, but also just assumed that the PD’s suspicions are correct.
Perhaps true, but to the same point you are doing that as well because he didn’t say the words you have their either. 
I made what I consider to be a reasonable read of what was said without additional context.
Let’s hope that no relevant data comes back as a result… imagine having to unplug these things and put them away every-time your mother in law visits… you know… just in case.
The meaning of what he said is identical to the meaning conveyed by what I said. You actually added meaning that was not present in what he said. Claiming that I’m doing the same thing because his wording was slightly different than mine is an exercise in pedantry.
If you consider…
“folks wanting to know if the hot cloud enabled mics they welcomed into the most intimate parts of their lives are a privacy burden beyond what they know or even a legal liability”
…to be a reasonable, let alone most likely correct interpretation of…
“And more fodder for the tin foil hat brigade.”
…then you and I have very, very different ideas of what is reasonable.
Yup. On this topic we do.
IMO you added WAY more meaning than I did.Perhaps you derived that from context not here in this thread - I have no idea.
So you say. And perhaps your correct, did he PM you on that?
Understatement of the year there.
I’d be more concerned over he Samsung TVs that are actually recording and sending all audio to a 3rd party for wake word analysis. This is a real privacy scare.
Echo’s allow you to delete recordings. And only what is recorded is after the wake word is sensed and if the blue lights are on it is sending the audio to amazon.
And it does have a mute button. Samsung smarttvs do not have an ability to temporarily turn off their audio processing.
Honestly, you’ve gone beyond ridiculous here.
“the tin foil hat brigade.” = “those who are paranoid about this issue”
“more fodder” = “and that this story will feed that paranoia”
Is this really that difficult? Tell me which parts of his statements in bold match up with…
“folks wanting to know if the hot cloud enabled mics they welcomed into the most intimate parts of their lives are a privacy burden beyond what they know or even a legal liability”
…and tell me again how I added way more meaning than you did.
And no, OP didn’t PM me. I’m just able to interpret simple English.
the smart water meter data seems more incriminating, than whatever Echo data they claim to have extracted.
How long before these gadgets are hacked to plant false evidence? yesterday? 10 years ago? Video is already susceptible to live tampering. Too many IoT sensors seem like soft targets waiting to be exploited.
I won’t be killing anybody, but I have no way to prove it.
@DParker
Chill out for a bit. I don’t have a beef with you or the other poster, no need to make a mountain out of a mole hill…
As my posts have already said - I think that privacy issues are important. I think there are plenty of examples of previously unknown privacy issues of a result of technology. I don’t think we should take Amazon at their word. I think that this incident is an opportunity press the issue and hopefully have confirmation that the way Amazon says the way it works is in fact the way it works. This will be a huge win for privacy.
I think there is an attempt in many situations attempt to disparage those that wish to protect and preserve privacy by using terms like ‘tin foil hat’ to describe those types of people - WHICH IS VERY COMMON EVEN IF THAT IS NOT WHAT OP INTENDED. That is how I read it. I believed that to be reasonable.
You can disagree with that, that goes to the very heart of interpretation. If you think your interpretation is infallible and the only reasonable way to interpret it - I can only point you to the long history of contradictory interpretations on very simple terms by our courts.
No reason to continue a needless debate.
Amazon claims it is more like 30 second loop stored on the device.
In this article, someone is quoted by NYT as saying it’s 60 seconds:
http://www.nytimes.com/2016/12/28/business/amazon-echo-murder-case-arkansas.html?_r=0
But it’s not Amazon claiming that, it’s a 3rd party.
In any event, 30, 60, 5 minutes, 10 minutes doesn’t seem like a meaningful difference to me - if it’s local and on a loop. If confirmed through legal process this is a good thing.
Make sure if someone starts stabbing you in your house you scream ‘Hey Alexa’ to get it up to amazon’s servers for permanent storage.
But Lynn Terwoerds, the executive director of the Executive Women’s Forum, which founded and sponsors the Voice Privacy Alliance, said in an email that the request for the information was built on a faulty premise.
She said the Echo is always listening for a “wake word” — Alexa, Amazon or another customizable term — and records only what is said after it has been activated. She said it has 60 seconds of recorded sound in its storage. “What this ‘always listening’ means is that the device is not eavesdropping and interpreting everything you’re saying,” Ms. Terwoerds wrote.
Once it detects the wake word, according to Amazon, the Echo starts streaming audio to the cloud, where it is secured until the customer permanently deletes it.
The case raises “serious privacy concerns with this kind of nonspecific warrant,” Ms. Terwoerds said, adding, “We have to fight against the myth of Echo listening in on our every word and sending that data to Amazon — it’s simply untrue.”
Alexa aside, the other IOT data even though seemingly unintrusive and would be scoffed at as ridiculous to be concerned about by many… turns out to be a issue as well… .
Interestingly, even in the Arkansas case, investigators are also using information from a smart water meter, alleging that an increase in water use in the middle of the night suggests a possible cleanup around the crime scene.
Imagine how much information/data from an extensive ST deployment (Barring Cameras and Microphones) alone to fit and sell a narrative crafted by investigators about alleged illegal acts in those ST homes.
Often when a family member is accused of being involved in murder you will hear the investigator on the case disparage the accused by saying “Their tears and emotional reaction at the news of the death seemed fake and forced” or alternatively “Their reaction wasn’t what you’d expect, there wasn’t an emotional outburst, etc”… any reaction or no reaction - you lose! There’s no winning.
Basically, everything - no matter what it is that you do or don’t do - no matter what the data is - will be twisted to fit the narrative and to paint you as the evil doer.
Yeah, and with my setup I can actually tell Alexa to record messages for me. They are stored and I can play then back when ready.
“Alexa. Please send my wife a message - I am being stabbed to death, I love you. Hug the kids for me. Goodbye.”
" … and don’t forget to flush the toilet about fifty times during the middle of the night because … well, because …"