This story was widely reported, but it's a little questionable. The slides are genuine, but they seem more like a prospectus for something Cox wants to do than something that they're actually doing. The presentation also never claims the "always listening" that people are concerned about, and instead just refers to "a data trail based on their conversations and online behavior" from "smart devices." The idea that this is smartphones listening to you pervasively is entirely something people have read into it, not something the slides say or even really suggest. I think most readers in the industry find it far more likely that they are describing reanalysis of consumer interactions with voice assistants (probably not even of the audio but of the transcript). That would presumably be the one in the cable boxes their parent company distributes, because access to that kind of data from other voice assistants seems difficult to negotiate and they do not claim to have it.

If you review the actual presentation (https://www.documentcloud.org/documents/25051283-cmg-pitch-d...), none of the claims in it are that remarkable. The whole kerfuffle seems to have come from a combination of the presentation actually being incredibly vague (probably intentionally to allow them to overstate the capability---this is a presentation for sales pitches) and some confirmation bias on the part of 404 Media, reading into it what they were looking for. But there's also a healthy amount of "news laundering," a lot of the articles cribbing off of 404 (like this techtimes one) actually make stronger claims than the original 404 piece does. It has a fair amount of weaseling that it's not clear how the slide deck should be interpreted, whether it's a real or speculative capability, etc.

If you've ever worked in enterprise software sales, you would be extremely wary of interpreting the slide deck the way a lot of these articles do. It reads like a lot of bluster.

I've never heard of them, so can't say if they're junk. But they're certainly gullible.

It's a small media company whose primary business is operating a handful of local newspapers and TV stations. They have no privileged access to mobile operating systems. If they really had implemented this scheme in actual apps (rather than just write it into a pitch deck), those apps would need to be asking for microphone permissions.

Note how there never was any follow-up showing that this really was happening. The story was only ever about that pitch deck. Compare that to e.g. the currently ongoing story about the dodgy things done by the Honey browser extension.

Discussed last year when they made their announcement. No one gave a believable mechanism for them being able to do this. The consensus was that they were blowing smoke.

https://arstechnica.com/gadgets/2023/12/no-a-marketing-firm-...

Ahh this one. Thanks for sharing, important point!

Basically this is just a random small-ish company trying to get new clients with a flashy feature. Ultimately they have to use the same data as everyone else, which I’m 99.99% sure doesn’t involve any intentional (much less “active”) recording by Google, Apple, or Meta. Maybe they have their own hardware partners that have networked microphones, maybe they’re really using incidental recordings from their “407 data partners”, or maybe it’s an empty promise - I sadly can’t read the original(ish) article https://www.404media.co/heres-the-pitch-deck-for-active-list...

It’s perhaps noteworthy that the intermediary source is the NYPost, which is most certainly junk! This story isn’t fake news, but it also isn’t presented in a honest way, IMHO

I had an argument about that exact story on lobste.rs a few months ago: https://lobste.rs/s/mf7guc/leak_facebook_partner_brags_about...

(I wouldn't exactly rate techtimes.com up there with NYT/Washington Post/etc.)