6 ms·

Alexa listens to more then the wake word. I've had this recommend products on Amazon while my wife and I were having a conversation at the dinner table. It also

Alexa listens to more then the wake word. I've had this recommend products on Amazon while my wife and I were having a conversation at the dinner table. It also recommended calling 911 while I was talking about fire. It's a spying device people willingly put in their homes for the convenience of a timer you can active with your voice. Edit: It could be that it activated mishearing "Alexa". I don't have hard evidence of mass spying. I think this wouldn't be hard to prove intercepting the data using something like Wireshark. Even if it's encrypted, you could tell by the data size. The recommending products while chatting with my wife anecdote happened multiple times tho, which convinced me to relocate the Alexa device in the garbage. It seems unlikely to me to not mine the voice data to generate ads, or to do law enforcement.

EGreg 2y ago
Just wait until people start buying Tesla’s optimus robots! Cheap and ubiquitous surveillance will be the norm in the future.
- margalabargala 2y ago
  Cheap and ubiquitous surveillance is the norm in the present.
  - yubblegum 2y ago
    Everyone carries a little snitch on them. Even if you opt out of using a mobile device, the chances of the person you are talking to having it on them is effectively 100%. And I am nearly certain that one way or another 'they' have voice biometrics on all of us. Thank god we live in a country with strong checks & balances...
  - sitkack 2y ago
    Every Tesla is constantly recording, so if anyone has been within 50 feet of a Tesla, they already have your picture.
    - margalabargala 2y ago
      Between cell phones, IOT doorbells, and incredibly inexpensive security cameras, even a world free from Tesla has cheap and ubiquitous surveillance. But you're correct, Tesla is yet another among the multitudes with a robust surveillance network
      - sitkack 2y ago
        Every starlink station (and probably) tesla, scoop up every mac address they ever see. This is one is unique in that it puts all that data into a single actor's hands.
        gkbrk 2y ago
        As opposed to every Android phone doing the same?
        sitkack 2y ago
        Continue
        margalabargala 2y ago
        Of course starlink stations scoop mac addresses. They are in this way equivalent to and on par with every other wifi router. A Tesla vehicle could also scoop up visible mac addresses, and is equally as capable of doing so as every other wifi-enabled device with closed source firmware. Android phones have been scooping wifi mac addresses, pairing them with GPS data, and sending that to google for at least 7 years: https://slate.com/technology/2018/06/how-google-uses-wi-fi-networks-to-figure-out-your-exact-location.html https://slate.com/technology/2018/06/how-google-uses-wi-fi-n... Apple probably has an equivalent system. Privacy-wise, Tesla is shitty but not extraordinarily shitty. Their surveillance capabilities do not differentiate them from among the multitudes. Let's assume maximum maliciousness. Assuming you don't own one, could Tesla track you particularly better than, say, Square? Or Google? Or Palantir? Or Comcast? Or any cell phone company? Or whomever it is that owns the cameras at each traffic light intersection?
- dingnuts 2y ago
  if people really think Musk is a Nazi, this would be like literally putting mindless order-following gestapo right in your house. Surveillance? Shit they could just kill you the moment you were discovered to be some undesirable. We're talking about a humanoid-ish robot, after all. If it can help you with the laundry it can bash your head in, too.
  - 52-6F-62 2y ago
    Pretty sure the “deletion” of undesirables is part of the plan, if a little bit further down the line. I thought they were waiting for ubiquitous AI micro drone technology. Maybe not. https://www.vcinfodocs.com/weapons-startups https://www.vcinfodocs.com/weapons-startups
  - solardev 2y ago
    On the bright side, at least I'll die with a clean shirt on.
    - wongarsu 2y ago
      It might even clean up the blood stains. Letting the blood dry would ruin the carpet and furniture
  - thih9 2y ago
    To be fair, we don’t need a new product, a tesla can do the same[1]. With AIs becoming more powerful and expanding to new areas, it makes even more sense to avoid businesses that are consistently user hostile. I wonder if anti tesla protests and related bad pr will contribute to increased consumer awareness around the topic. [1]: e.g. it can crash into a will e coyote style wall on autopilot: https://www.youtube.com/watch?v=IQJL3htsDyQ&t=899s https://www.youtube.com/watch?v=IQJL3htsDyQ&t=899s
    - EGreg 2y ago
      If there’s one thing about AI, it’s that you cannot avoid it. The idea that individuals can just “opt out” of plastic, sugar, artificial ingredients, factory farms, social media and all the other negative extrnalities the corporations push on us is a fantasy that governments and industry push on individuals to keep us distracted: https://magarshak.com/blog/?p=362 https://magarshak.com/blog/?p=362 On HN, people hate on Web3 because of its limited upside. But really look at the downside dynamics of a technology! With Web3, you can only ever lose what you voluntarily put in (at great effort and slippage LOL). So that caps the downside. Millions of people who never got a crypto wallet and never sent their money to some shady exchange never lost a penny. Now compare that to AI. No matter what you do, no matter how far you try to avoid it millions will lose their jobs, get denied loans, be surveiled, possibly arrested for precrime, micromanaged and controlled, practically enslaved in order to survive and reproduce etc. It won’t even work to retreat into gated communities or grandfathered human-verified accounts because defectors will run bots in their accounts and their neuralink cyborg hookups and meta glasses, to gain an advantage and approach at least some of the advantages of the bots. Not to mention of course that the economic power and efficiency of botless communities will be laughably uncompetitive. You won’t even be able to move away anywhere to escape it. You can see an early preview of that with the story of Ted Kazinsky — the unabomber (google it). While the guy was clearly a disturbed maniac who sent explosives to people, as a mathematician following things to its logical conclusion he did sort of predict what will happen to everyone when technology reaches a certain point. AI just makes it so that you can’t escape. If HN cared about AI unlimited downsides like it cared about Web’s lack of large upsides, the sentiment here would be very different. But the time has not come yet. Set an alarm to check back on this comment in exactly 7 years.
- r00fus 2y ago
  Telsa has an uncertain future, much less capability to actually pull off robots.
- taneq 2y ago
  That’s how Skynet wins!
ranger_danger 2y ago
Extraordinary claims require extraordinary evidence.
- ethbr1 2y ago
  Should not the burden of proof be on Amazon to prove it's not always recording? In 2025, it feels like we're 5 to 10 years past the time a consumer should default to assuming their cloud-connected device isn't extracting the maximum possible revenue from them. Assume all companies are amoral, and you'll never be disappointed.
  - bgirard 2y ago
    > Should not the burden of proof be on Amazon to prove it's not always recording? What's a satisfactory burden of proof for Amazon to meet here? There's only so much you can do to prove that something is NOT happening.
    - whateveracct 2y ago
      Okay..but then why should I trust that Alexa isn't listening? That's clearly a pretty valuable thing for Amazon to provide to their customers. Is it impossible? If it is..then yeah people should just light these things on fire or have a hard switch on them at least.
    - yeahsure 2y ago
      Well, for starters, I guess keeping the "do not send voice recordings" toggle would be a good idea.
    - fardo 2y ago
      They have a lot of ways they could’ve built trust without a full negative burden: which of them, if any, are they doing? Open sourcing of their watch word and recording features specifically, so people can self-verify it does what it says and that it’s not doing sketchy things? Hardware lights such that any record functionality past the watch words is visible and verifiable by the end user and it can’t record when not lit? Local streaming and auditable downloads of the last N hours of input as heard by amazon after watchwords, so you can check for misrecordings and also compare “intended usage” times to observed times, such that you can see that you and Amazon get the same stuff? If you really wanna go all out, putting in their TOS protections like explicit no-train permissions on passing utterances without intent, or adding an SLA into their subscription to refund subscription and legal costs and to provide explicit legal cause of action, if they were recording when they said they weren’t? If you explicitly want to promote trust, there are actually a ton of ways to do it, one of them isn’t “remove even more of your existing privacy guardrails”.
mmaunder 2y ago
I have Alexa and a few Siri devices next to me and I just said a bunch of phrases indicating fire, choking, that we should call 911 etc and nothing triggered. So yeah - this is just internet bullshit until proven otherwise.
- reverendsteveii 2y ago
  it both is and isn't secret bullshit. there's no evidence that there's a list of secret keywords Lord Bezos is listening for, but there's plenty of evidence that these devices active unintentionally all the time and that those unintentional activations lead to you being recorded and that recording being sent off into the cloud https://moniotrlab.khoury.northeastern.edu/publications/smart-speakers-study-pets20/ https://moniotrlab.khoury.northeastern.edu/publications/smar...
  - ranger_danger 2y ago
    I don't think it's any secret that the device can unintentionally activate in certain circumstances (and whether or not that's due to it thinking it heard its name is another debate)... but my problem with OP's statement is that they seem to frame it as if it's intentionally and maliciously listening more often than it should, and I just don't see any evidence to support that claim.
    - walterbell 2y ago
      Alexa option to "recognize sounds" (e.g. baby crying, fire alarm, appliance beeping) might increase risk of false positives.
    - reverendsteveii 2y ago
      What I'm saying is that intentionality doesn't have to be relevant to this discussion. All you need to do in order to be maliciously spying on someone, given that you have this bug in the first place, is to 1) not fix the bug 2) quietly remove the option to opt out of remote processing and then all of a sudden you've got a situation where of course no one is actively spying because We Would Never(tm)(c)(r) but there's a really reliable pipeline by which recordings of me talking to me family in my home end up on a remote server somewhere where they're used to train AI and maybe even automatically scanned for certain keywords that might indicate that I'm some sort of troublemaker and need flagged for additional "attention". It's a plausibly-deniable panopticon. In fact having it activate by purposefully unremediated mistake rather than by keyword makes it a better spy. You can discover a list of keywords and avoid them but ambient noise causing the device to randomly sample and exfiltrate recordings means you can never know when you're being recorded and thus have no choice but to always act like you're being recorded, just in case.
BurningFrog 2y ago
This is just an anecdote from an anonymous internet account. I don't necessarily trust Amazon, but I trust anonymous internet accounts far less. If Amazon does spy this blatantly, it should be easy to verify and reproduce empirically. Has anyone serious done that?
- bentt 2y ago
  That's always been my perspective. The incentive for busting Amazon on this is so high, if it were provable then someone would have done it and the press would love to share that. My point is it no longer matters. We're out.
ethbr1 2y ago
As always, there's a hilariously apropos "imagine if technology were this bad" that turns out to just be reality. Amazon: We're listening. https://m.youtube.com/watch?v=JCQhygMdFr4&t=57s https://m.youtube.com/watch?v=JCQhygMdFr4&t=57s
- sitkack 2y ago
  We see you. We hear you. https://www.youtube.com/watch?v=PQw0-AkgQGM https://www.youtube.com/watch?v=PQw0-AkgQGM
janalsncm 2y ago
And if that’s really all you’re using it for there’s no reason it needs Internet at all. All of the intelligence can be run on-device.
reverendsteveii 2y ago
https://moniotrlab.khoury.northeastern.edu/publications/smart-speakers-study-pets20/ https://moniotrlab.khoury.northeastern.edu/publications/smar... Let me float an idea based on this study: it doesn't need to listen for more than the wake word (and some variations of same) in order to activate pretty regularly. The study indicates that just watching TV in a room w an echo device will cause it to wake 1-4 times per hour with half of all wakes resulting in at least 4 seconds of recording and virtually all recordings being sent to the cloud for processing. Even absent any "secret wake words" the device activates regularly enough that it will occasionally react to things you say in conversation as though it's secretly listening. Also, just thinking like someone who is simultaneously evil and competent, if I were building a device like this that listened for secret keywords I wouldn't have it announce the fact that it heard one.
- walterbell 2y ago
  > wouldn't have it announce the fact that it heard one Alexa has an option to play a sound for "Start of Request", i.e. wake word heard. There is a ring light on the device, blue when it has heard a wake word, red when the hardware mute button is pressed.
tartoran 2y ago
Corpos can't resist not to spy when it's at their fingertips, too irresistible to them, they just can't help it. That's why we should take our privacy back and offer no benefit of the doubt.
- walterbell 2y ago
  IoT WiFi baseband firmware is also an attack surface.
globular-toast 2y ago
The fact you have two such anecdotes means you kept the spying device in your home.
cryptoegorophy 2y ago
Yeah, what’s the best thing that can happen? It is just an Alexa device, and the worst? It leaks your data and spies on you. No thanks.
frereubu 2y ago
I'm not sure whether it's listening to more than the wake word, but I've seen Siri wake up quite often when I very definitely haven't said anything approaching "Siri", and see it occasionally on other people's devices too. I remember listening to a BBC podcast in the car once and there was once piece of audio from it that would reliably activate Siri. I was a bit nonplussed by it and rewound it four or five times to check, which it did every time. I think accidental activation is a much more likely explanation, which is still dreadful from a privacy perspective.

EGreg 2y ago

Just wait until people start buying Tesla’s optimus robots! Cheap and ubiquitous surveillance will be the norm in the future.

margalabargala 2y ago
Cheap and ubiquitous surveillance is the norm in the present.
- yubblegum 2y ago
  Everyone carries a little snitch on them. Even if you opt out of using a mobile device, the chances of the person you are talking to having it on them is effectively 100%. And I am nearly certain that one way or another 'they' have voice biometrics on all of us. Thank god we live in a country with strong checks & balances...
- sitkack 2y ago
  Every Tesla is constantly recording, so if anyone has been within 50 feet of a Tesla, they already have your picture.
  - margalabargala 2y ago
    Between cell phones, IOT doorbells, and incredibly inexpensive security cameras, even a world free from Tesla has cheap and ubiquitous surveillance. But you're correct, Tesla is yet another among the multitudes with a robust surveillance network
    - sitkack 2y ago
      Every starlink station (and probably) tesla, scoop up every mac address they ever see. This is one is unique in that it puts all that data into a single actor's hands.
      - gkbrk 2y ago
        As opposed to every Android phone doing the same?
        sitkack 2y ago
        Continue
      - margalabargala 2y ago
        Of course starlink stations scoop mac addresses. They are in this way equivalent to and on par with every other wifi router. A Tesla vehicle could also scoop up visible mac addresses, and is equally as capable of doing so as every other wifi-enabled device with closed source firmware. Android phones have been scooping wifi mac addresses, pairing them with GPS data, and sending that to google for at least 7 years: https://slate.com/technology/2018/06/how-google-uses-wi-fi-networks-to-figure-out-your-exact-location.html https://slate.com/technology/2018/06/how-google-uses-wi-fi-n... Apple probably has an equivalent system. Privacy-wise, Tesla is shitty but not extraordinarily shitty. Their surveillance capabilities do not differentiate them from among the multitudes. Let's assume maximum maliciousness. Assuming you don't own one, could Tesla track you particularly better than, say, Square? Or Google? Or Palantir? Or Comcast? Or any cell phone company? Or whomever it is that owns the cameras at each traffic light intersection?
dingnuts 2y ago
if people really think Musk is a Nazi, this would be like literally putting mindless order-following gestapo right in your house. Surveillance? Shit they could just kill you the moment you were discovered to be some undesirable. We're talking about a humanoid-ish robot, after all. If it can help you with the laundry it can bash your head in, too.
- 52-6F-62 2y ago
  Pretty sure the “deletion” of undesirables is part of the plan, if a little bit further down the line. I thought they were waiting for ubiquitous AI micro drone technology. Maybe not. https://www.vcinfodocs.com/weapons-startups https://www.vcinfodocs.com/weapons-startups
- solardev 2y ago
  On the bright side, at least I'll die with a clean shirt on.
  - wongarsu 2y ago
    It might even clean up the blood stains. Letting the blood dry would ruin the carpet and furniture
- thih9 2y ago
  To be fair, we don’t need a new product, a tesla can do the same[1]. With AIs becoming more powerful and expanding to new areas, it makes even more sense to avoid businesses that are consistently user hostile. I wonder if anti tesla protests and related bad pr will contribute to increased consumer awareness around the topic. [1]: e.g. it can crash into a will e coyote style wall on autopilot: https://www.youtube.com/watch?v=IQJL3htsDyQ&t=899s https://www.youtube.com/watch?v=IQJL3htsDyQ&t=899s
  - EGreg 2y ago
    If there’s one thing about AI, it’s that you cannot avoid it. The idea that individuals can just “opt out” of plastic, sugar, artificial ingredients, factory farms, social media and all the other negative extrnalities the corporations push on us is a fantasy that governments and industry push on individuals to keep us distracted: https://magarshak.com/blog/?p=362 https://magarshak.com/blog/?p=362 On HN, people hate on Web3 because of its limited upside. But really look at the downside dynamics of a technology! With Web3, you can only ever lose what you voluntarily put in (at great effort and slippage LOL). So that caps the downside. Millions of people who never got a crypto wallet and never sent their money to some shady exchange never lost a penny. Now compare that to AI. No matter what you do, no matter how far you try to avoid it millions will lose their jobs, get denied loans, be surveiled, possibly arrested for precrime, micromanaged and controlled, practically enslaved in order to survive and reproduce etc. It won’t even work to retreat into gated communities or grandfathered human-verified accounts because defectors will run bots in their accounts and their neuralink cyborg hookups and meta glasses, to gain an advantage and approach at least some of the advantages of the bots. Not to mention of course that the economic power and efficiency of botless communities will be laughably uncompetitive. You won’t even be able to move away anywhere to escape it. You can see an early preview of that with the story of Ted Kazinsky — the unabomber (google it). While the guy was clearly a disturbed maniac who sent explosives to people, as a mathematician following things to its logical conclusion he did sort of predict what will happen to everyone when technology reaches a certain point. AI just makes it so that you can’t escape. If HN cared about AI unlimited downsides like it cared about Web’s lack of large upsides, the sentiment here would be very different. But the time has not come yet. Set an alarm to check back on this comment in exactly 7 years.
r00fus 2y ago
Telsa has an uncertain future, much less capability to actually pull off robots.
taneq 2y ago
That’s how Skynet wins!

ranger_danger 2y ago

Extraordinary claims require extraordinary evidence.

ethbr1 2y ago
Should not the burden of proof be on Amazon to prove it's not always recording? In 2025, it feels like we're 5 to 10 years past the time a consumer should default to assuming their cloud-connected device isn't extracting the maximum possible revenue from them. Assume all companies are amoral, and you'll never be disappointed.
- bgirard 2y ago
  > Should not the burden of proof be on Amazon to prove it's not always recording? What's a satisfactory burden of proof for Amazon to meet here? There's only so much you can do to prove that something is NOT happening.
  - whateveracct 2y ago
    Okay..but then why should I trust that Alexa isn't listening? That's clearly a pretty valuable thing for Amazon to provide to their customers. Is it impossible? If it is..then yeah people should just light these things on fire or have a hard switch on them at least.
  - yeahsure 2y ago
    Well, for starters, I guess keeping the "do not send voice recordings" toggle would be a good idea.
  - fardo 2y ago
    They have a lot of ways they could’ve built trust without a full negative burden: which of them, if any, are they doing? Open sourcing of their watch word and recording features specifically, so people can self-verify it does what it says and that it’s not doing sketchy things? Hardware lights such that any record functionality past the watch words is visible and verifiable by the end user and it can’t record when not lit? Local streaming and auditable downloads of the last N hours of input as heard by amazon after watchwords, so you can check for misrecordings and also compare “intended usage” times to observed times, such that you can see that you and Amazon get the same stuff? If you really wanna go all out, putting in their TOS protections like explicit no-train permissions on passing utterances without intent, or adding an SLA into their subscription to refund subscription and legal costs and to provide explicit legal cause of action, if they were recording when they said they weren’t? If you explicitly want to promote trust, there are actually a ton of ways to do it, one of them isn’t “remove even more of your existing privacy guardrails”.

mmaunder 2y ago

I have Alexa and a few Siri devices next to me and I just said a bunch of phrases indicating fire, choking, that we should call 911 etc and nothing triggered. So yeah - this is just internet bullshit until proven otherwise.

reverendsteveii 2y ago
it both is and isn't secret bullshit. there's no evidence that there's a list of secret keywords Lord Bezos is listening for, but there's plenty of evidence that these devices active unintentionally all the time and that those unintentional activations lead to you being recorded and that recording being sent off into the cloud https://moniotrlab.khoury.northeastern.edu/publications/smart-speakers-study-pets20/ https://moniotrlab.khoury.northeastern.edu/publications/smar...
- ranger_danger 2y ago
  I don't think it's any secret that the device can unintentionally activate in certain circumstances (and whether or not that's due to it thinking it heard its name is another debate)... but my problem with OP's statement is that they seem to frame it as if it's intentionally and maliciously listening more often than it should, and I just don't see any evidence to support that claim.
  - walterbell 2y ago
    Alexa option to "recognize sounds" (e.g. baby crying, fire alarm, appliance beeping) might increase risk of false positives.
  - reverendsteveii 2y ago
    What I'm saying is that intentionality doesn't have to be relevant to this discussion. All you need to do in order to be maliciously spying on someone, given that you have this bug in the first place, is to 1) not fix the bug 2) quietly remove the option to opt out of remote processing and then all of a sudden you've got a situation where of course no one is actively spying because We Would Never(tm)(c)(r) but there's a really reliable pipeline by which recordings of me talking to me family in my home end up on a remote server somewhere where they're used to train AI and maybe even automatically scanned for certain keywords that might indicate that I'm some sort of troublemaker and need flagged for additional "attention". It's a plausibly-deniable panopticon. In fact having it activate by purposefully unremediated mistake rather than by keyword makes it a better spy. You can discover a list of keywords and avoid them but ambient noise causing the device to randomly sample and exfiltrate recordings means you can never know when you're being recorded and thus have no choice but to always act like you're being recorded, just in case.

BurningFrog 2y ago

This is just an anecdote from an anonymous internet account. I don't necessarily trust Amazon, but I trust anonymous internet accounts far less. If Amazon does spy this blatantly, it should be easy to verify and reproduce empirically. Has anyone serious done that?

bentt 2y ago
That's always been my perspective. The incentive for busting Amazon on this is so high, if it were provable then someone would have done it and the press would love to share that. My point is it no longer matters. We're out.

ethbr1 2y ago

As always, there's a hilariously apropos "imagine if technology were this bad" that turns out to just be reality. Amazon: We're listening. https://m.youtube.com/watch?v=JCQhygMdFr4&t=57s https://m.youtube.com/watch?v=JCQhygMdFr4&t=57s

sitkack 2y ago
We see you. We hear you. https://www.youtube.com/watch?v=PQw0-AkgQGM https://www.youtube.com/watch?v=PQw0-AkgQGM

janalsncm 2y ago

And if that’s really all you’re using it for there’s no reason it needs Internet at all. All of the intelligence can be run on-device.

reverendsteveii 2y ago

https://moniotrlab.khoury.northeastern.edu/publications/smart-speakers-study-pets20/ https://moniotrlab.khoury.northeastern.edu/publications/smar... Let me float an idea based on this study: it doesn't need to listen for more than the wake word (and some variations of same) in order to activate pretty regularly. The study indicates that just watching TV in a room w an echo device will cause it to wake 1-4 times per hour with half of all wakes resulting in at least 4 seconds of recording and virtually all recordings being sent to the cloud for processing. Even absent any "secret wake words" the device activates regularly enough that it will occasionally react to things you say in conversation as though it's secretly listening. Also, just thinking like someone who is simultaneously evil and competent, if I were building a device like this that listened for secret keywords I wouldn't have it announce the fact that it heard one.

walterbell 2y ago
> wouldn't have it announce the fact that it heard one Alexa has an option to play a sound for "Start of Request", i.e. wake word heard. There is a ring light on the device, blue when it has heard a wake word, red when the hardware mute button is pressed.

tartoran 2y ago

Corpos can't resist not to spy when it's at their fingertips, too irresistible to them, they just can't help it. That's why we should take our privacy back and offer no benefit of the doubt.

walterbell 2y ago
IoT WiFi baseband firmware is also an attack surface.

globular-toast 2y ago

The fact you have two such anecdotes means you kept the spying device in your home.

cryptoegorophy 2y ago

Yeah, what’s the best thing that can happen? It is just an Alexa device, and the worst? It leaks your data and spies on you. No thanks.

frereubu 2y ago

I'm not sure whether it's listening to more than the wake word, but I've seen Siri wake up quite often when I very definitely haven't said anything approaching "Siri", and see it occasionally on other people's devices too. I remember listening to a BBC podcast in the car once and there was once piece of audio from it that would reliably activate Siri. I was a bit nonplussed by it and rewound it four or five times to check, which it did every time. I think accidental activation is a much more likely explanation, which is still dreadful from a privacy perspective.