Also, this person is, well, for a lack of a better term, a porn actress. So this tweet might be a mix of bias, leading to the belief their phone is listening, but maybe also a bit of advertisement - which, don’t get me wrong, is great. She did some stuff with Owen Gray, good stuff. Check them out, support their work.
That being said, it is certainly technically feasible. I’m sure there have been several, individual cases, and we will probably see new, “innovative” ways of companies spying on us.
Dude, there’s more, but this is the first one that came up.
A leaked pitch deck from CMG Local Solutions, a subsidiary of Cox Media Group (CMG), details a method it calls “active listening.” This method uses AI to combine voice data with online behavioral data to deliver hyper-targeted advertising.
The deck, obtained by 404 Media, states, “Advertisers can pair this voice-data with behavioral data to target in-market consumers.” It goes on to say that the technology can identify “ready-to-buy” consumers and create ad lists based on their spoken intentions.
As I said, I’m sure companies will try. But what you’re looking at there is a pitch deck presented to someone so they’d cough up money. That does in no way imply that this has been widely deployed without requiring user consent and therefore there are apps out there en masse, listening to smartphone users in a form of clandestine operation. It’s basically the same thing as the patent for the old greentext Mountain Dew commercial meme at this point:
In fact, the article clearly states that the data Cox Media Group uses comes from apps where users have agreed to grant the application permission to use their voice data, and that Cox Media Group was subsequently removed from the Google Ads program (a precaution for Google to save face).
I used to think they didn’t listen, and I’d cite the same sort of studies. But honestly…I’ve had a few times in the last several months where a conversation I had with my GF triggered ads on our phones, when neither of us had done an internet search of that particular subject or any that were tangential. A couple times, now, I’ve been spooked. Could be some other data we put out there, or something we clicked, but I racked my brain and couldn’t think of any.
Maybe our phones aren’t listening to us all the time. But it’s already pretty fucking bad when lived experience makes us wonder whether they really are.
[The] lawsuit that alleged Apple was infringing on its users’ privacy by capturing conversations overheard by its Siri voice assistant without consent, passing the recordings to third-party quality control contractors. Apple offered a formal apology and pledged it would no longer retain user recordings, but pushed back against additional allegations that it allowed advertisers to target consumers based on Siri recording data. In January 2025, the company agreed to pay $95 million out to impacted users to settle the case.
So apple was doing this and apologized for it.
As an aside, if you have a voice activated assistant like Siri enabled, it is constantly listening for the voice activation command.
Not accessing the microphone doesn’t necessarily mean you aren’t listening in.
I read an article at one point about how the gyrometer can be used as a microphone as well. Not as good and not all phones but.
Yeah while absence of evidence is not evidence of absence and I too sometimes think that a phone must have been listening in… idk. Might just be the frequency illusion. Especially if she’s a porn actress, I can imagine a billion things in her interactions on the phone without it listening in would also probably lead to similar adverts.
I happen to work in machine learning. You are most likely referring to the Stanford Gyrophone paper. Given that the sampling frequency of the gyroscope sensor on typical smartphones is extremely limited, you can only get very low frequency content (Nyquist).
It wouldn’t be possible for any human to process or understand the recorded signals, so the researchers trained a machine learning model on the recorded samples, with a very limited vocabulary consisting of only the digits from 0 to 9 and “oh”.
If the model was not trained on the particular speaker (requiring annotated training data for that particular speaker, which would be almost impossible to get in the assumed scenario), the recognition rate was 26%. For a vocabulary of 11 words.
It’s a nice proof of concept, and doubly so if tge CIA considers you a target, but otherwise it’s not happening.
You’re full of shit and must work for ad companies. Because they damn sure are listening. I seen proof. Especially when you phone calls out, “I didn’t get that” when your not even using the phone.
Also I get ads just by having a conversation about a subject they me and my wife discussed but never looked up. They do turn on those mics and use what they hear.
This is actually not a thing. There is no evidence that smartphones constantly listen to people.
https://www.researchgate.net/publication/334214258_Is_My_Phone_Listening_in_On_the_Feasibility_and_Detectability_of_Mobile_Eavesdropping
https://news.northeastern.edu/2018/07/06/is-your-smartphone-spying-on-you/
Also, this person is, well, for a lack of a better term, a porn actress. So this tweet might be a mix of bias, leading to the belief their phone is listening, but maybe also a bit of advertisement - which, don’t get me wrong, is great. She did some stuff with Owen Gray, good stuff. Check them out, support their work.
That being said, it is certainly technically feasible. I’m sure there have been several, individual cases, and we will probably see new, “innovative” ways of companies spying on us.
Dude, there’s more, but this is the first one that came up.
https://www.newsweek.com/phone-voice-assistants-active-listening-consent-targeted-ads-1949251
As I said, I’m sure companies will try. But what you’re looking at there is a pitch deck presented to someone so they’d cough up money. That does in no way imply that this has been widely deployed without requiring user consent and therefore there are apps out there en masse, listening to smartphone users in a form of clandestine operation. It’s basically the same thing as the patent for the old greentext Mountain Dew commercial meme at this point:
https://www.snopes.com/fact-check/sony-patent-mcdonalds/
In fact, the article clearly states that the data Cox Media Group uses comes from apps where users have agreed to grant the application permission to use their voice data, and that Cox Media Group was subsequently removed from the Google Ads program (a precaution for Google to save face).
I used to think they didn’t listen, and I’d cite the same sort of studies. But honestly…I’ve had a few times in the last several months where a conversation I had with my GF triggered ads on our phones, when neither of us had done an internet search of that particular subject or any that were tangential. A couple times, now, I’ve been spooked. Could be some other data we put out there, or something we clicked, but I racked my brain and couldn’t think of any.
Maybe our phones aren’t listening to us all the time. But it’s already pretty fucking bad when lived experience makes us wonder whether they really are.
yeah it’s not an Alexa
There is this recently settled lawsuit:
So apple was doing this and apologized for it.
As an aside, if you have a voice activated assistant like Siri enabled, it is constantly listening for the voice activation command.
And why did you cite some pretty old studies?
Not accessing the microphone doesn’t necessarily mean you aren’t listening in.
I read an article at one point about how the gyrometer can be used as a microphone as well. Not as good and not all phones but.
Yeah while absence of evidence is not evidence of absence and I too sometimes think that a phone must have been listening in… idk. Might just be the frequency illusion. Especially if she’s a porn actress, I can imagine a billion things in her interactions on the phone without it listening in would also probably lead to similar adverts.
I happen to work in machine learning. You are most likely referring to the Stanford Gyrophone paper. Given that the sampling frequency of the gyroscope sensor on typical smartphones is extremely limited, you can only get very low frequency content (Nyquist).
It wouldn’t be possible for any human to process or understand the recorded signals, so the researchers trained a machine learning model on the recorded samples, with a very limited vocabulary consisting of only the digits from 0 to 9 and “oh”.
If the model was not trained on the particular speaker (requiring annotated training data for that particular speaker, which would be almost impossible to get in the assumed scenario), the recognition rate was 26%. For a vocabulary of 11 words.
It’s a nice proof of concept, and doubly so if tge CIA considers you a target, but otherwise it’s not happening.
You’re full of shit and must work for ad companies. Because they damn sure are listening. I seen proof. Especially when you phone calls out, “I didn’t get that” when your not even using the phone.
Also I get ads just by having a conversation about a subject they me and my wife discussed but never looked up. They do turn on those mics and use what they hear.