Company Brags About Using Smart Device Microphone Audio to Target Ads

[jack_christie]

MindSift, a small New Hampshire-based company, is part of a new push that aims to target ads by listening to peoples’ everyday conversations through microphones in their smart devices, according to a review of recently deleted sections from MindSift’s website and comments made on a podcast

Company Brags About Using Smart Device Microphone Audio to Target Ads on Their Podcast

MindSift is the tiny company that says it’s using device microphone data to target ads.

www.404media.co

Should be more to come on this story...

 

[93Flareside]

Well at least it's no longer "oh you're just imagining it" type of gaslighting.

I'll write something to somebody in Telegram or WhatsApp and then a few hours later see ads for something I had written about. So fuck all of them.

 

[jack_christie]

Well at least it's no longer "oh you're just imagining it" type of gaslighting.

I'll write something to somebody in Telegram or WhatsApp and then a few hours later see ads for something I had written about. So fuck all of them.

All the more reason to use signal or matrix.

The 404media podcast has a discussion about the audio ads.

 

[Dr_Grip]

At least WhatsApp is end to end encyrpted, actually using the same stack as Signal. So believing your data is mined for ad targeting is paranoia.
Google Mail is another topic...

EDIT: Also I strongly doubt that this works on iOS (or even recent versions of Android).

 

[eizbaer]

I love how because it’s a company that’s bragging about this (and not just some dude online) and it’s so nicely in line with everyone’s prejudices, it somehow gets taken seriously…

 

[jack_christie]

I love how because it’s a company that’s bragging about this (and not just some dude online) and it’s so nicely in line with everyone’s prejudices, it somehow gets taken seriously…

The data was for sale on Cox media group

https://archive.ph/m1tbw

 

[Dr_Grip]

I still call this a fake. If this would be out there and deployed at scale
A) me or @_Jules_ would have heard a sales pitch from out media agencies about it
B) some security researchers would have found the data leaving the device (just as they found out about TikTok basically always listening)
C) Data use for devices would be significant since the data processing would have to happen on the server side, requiring a constant audio stream being sent out.

 

[Dr_Grip]

Also @jack_christie media (=advertising space and targeting) sales is the closest to organized crime that you'll find among "legit" businesses, so anything found in a media agency pitch deck may just be a bunch of lies.
As long as the big media agencies (dare I call them syndicates) like GroupM or Dentsu don't offer it I have doubts. This would be a massive competitive advantage so you'd expect the big dogs to have it.

 

[eizbaer]

anything found in a media agency pitch deck may just be a bunch of lies.

I’d also argue the same is true for the entire tech space… in that particular context it should be doubly true that just because someone is selling something, doesn’t mean that it actually exists (as advertised or even at all). There’s good reason for the emergence of the term vaporware…

 

[Blind_Io]

At least WhatsApp is end to end encyrpted, actually using the same stack as Signal. So believing your data is mined for ad targeting is paranoia.
Google Mail is another topic...

EDIT: Also I strongly doubt that this works on iOS (or even recent versions of Android).

Assuming it's not exploiting something in the on screen keyboard before the message is ever transmitted.

 

[Dr_Grip]

Assuming it's not exploiting something in the on screen keyboard before the message is ever transmitted.

Well, we know google does that (partly), pretty sure they would not let key competitor Meta in.

Mind you, Telegram for example is not encrypted by default and may well mine your data. Google acknowledges doing so in their terms of use, does Meta for timeline posts. But WhatsApp can be considered safe.

 

[Dr_Grip]

I just wanted to add/emphasize that I am not downplaying online privacy risks. I just feel that the border of paranoia and fatalism of "they read/hear anything anyways" does not help in making informed decisions, for example accepting that (eventhough it's counterintuitive) Zuckerberg-owned WhatsApp actually is safer than Telegram.