Perusing a paper paper for once I saw this advert for WhatsApp.
No I can believe the content of your message can not be read, but by using it, your address book is theirs, your messages sent/received are logged and you will be tracked wherever you are - and whatever you are buying.
That’s what they really want.
Go #Signal - it makes sense
For those who are interested I have made a formal complaint to the ASA -
The ad appeared in The Times on Saturday 12 July and the photo of the ad is in my original text. I will post any responses I get from them (other than the acknowledgement). Feel free to do it yourselves. The lying bar stewards.
@Wen And the most pervert thing is, IMHO, that the #whatsapp encryption is indeed strong, preventing the *service* from leaking. It is sort of a "safe pipe"...
... But since #Meta has access to your keyboard, microphone, camera, contacts, etc, they can anyway intercept anything you type, do and say even before it is sent through this "safe pipe". Same thing on the receiver side!
Even worse: why the hell would they advertise if it would not see a clear return on investment, either financial or via the exploitation of our private data to feed their #IA and profiling algorithms?
I don't get how so many people don't question this!
@gaufff And the app is closed source, so you don't actually know what it is doing. It might decrypt your message and send it somewhere else. Illegality doesn't bother Meta, we know that.
All of this is hopeless though, because everyone is using it. Eventually when Meta has taken over the entire communications infrastructure, only then will someone say, was that a good idea?
#signal #whatsapp #privacy #security #crime #organisedcrime
@Okuna @Wen I'm not *the* expert at this, maybe others will be able to elaborate on this, but I do not trust this at all.
When you give them access (mic for example), nothing states that you give them access "for this app only", right? To me, it means that they have access to those "sources" all the time.
The many many stories about people getting ads for stuff they just discussed while their phone was around (but not actively used) tends to confirm this.
To me, this is in no way paranoia. But such apps being closed source, we can never know what they do for sure.
@gaufff @Okuna You can frequently do this on a per app/per active app basis, but I am sure many don’t.
At the same time, I am dubious on the (multiple) claims of listening in for ads. Two reasons, 1) their use would leak 2) some will be coincidence but some will be tracking individuals who are linked already - so if Fred is interested in bondage aids, Wilma, in their address book might get the same ads (simplified links obviously )
@Okuna @Wen I appreciate the feedback, thank you. And as I said, my overall knowledge is quite OK but my detailed technical knowledge is very limited. Especially on mobile devices (#iPhone , #Android , alternatives like #GrapheneOS , ...).
Still: is it still fair to say that my trust could indeed be higher if source code of these apps (apps developed by the GAFAMs and run on devices sold by those GAFAMs) would be open and available to all for review and contribution? Are apps really sandboxes on all those #OS ?
Without that, what prevents them from doing exactly what they want with our #privacy
@Wen @helma and their AI cán actually read it. 👀
I've posted something about it for easy reference:
https://roelant.net/en/2025/meta-ai-has-access-to-your-encrypted-whatsapp-messages/
@roelant @Wen @helma Meta AI messages on WhatsApp use E2EE, but obviously the key is specifically shared with Meta, so that they can generate the response and send it back via a message.
It's clearly mentioned in the WhatsApp Help Center (https://faq.whatsapp.com/1002544104126998):
"The most important thing to know is that your personal messages with friends and family are off limits. AIs can read what is shared with them. [...] Talking to an AI provided by Meta doesn't link you personal WhatsApp account information on Facebook, Instagram, or any other apps provided by Meta." and About Meta AI (https://faq.whatsapp.com/2257017191175152):
"When you choose to use these features, Meta receives your prompts, the messages you share, and feedback to deliver relevant responses directly to you.
Only people that mention @/Meta AI, or that people choose to share with Meta AI, can be read by Meta. Meta can't read any other messages in your personal chats."
Also sourcing their white paper for more technical details: https://scontent.fclj2-1.fna.fbcdn.net/v/t39.8562-6/456045578_868167208055607_7997729792527463495_n.pdf?_nc_cat=107&ccb=1-7&_nc_sid=b8d81d&_nc_ohc=Zw7UZPd_vngQ7kNvwE7lwYu&_nc_oc=Adlbx217y4C7z4SM9DtZBeCjNwCNO7ZqeCxWATsh0TkbCWLkQ3qJj1gBY0pitk6q9bRDHV89-1z8jlkkrUuLGR5I&_nc_zt=14&_nc_ht=scontent.fclj2-1.fna&_nc_gid=aEv7_QsE_aLJzeCbH9qJiA&oh=00_AfT3kmxw0WVkcE2CI7j5cF7Wiyvd_Rz8-dXOKFvfKhwttA&oe=68782163
@roelant @Wen @helma And as for the article you cited, The Guardian has received a response from a Meta spokesperson:
"Meta AI is trained on a combination of licensed and publicly available datasets, not on the phone numbers people use to register for WhatsApp or their private conversations,” a spokesperson said. “A quick online search shows the phone number mistakenly provided by Meta AI is both publicly available and shares the same first five digits as the TransPennine Express customer service number.”
So, nothing private was shared. The AI hallucinated a phone number similar to the official one, or the supposedly 'private' phone number was actually part of a public dataset that Meta AI used for the training data. Again, no personal information was shared from WhatsApp.
@alextecplayz @roelant @helma And from the same FAQ
'What you send to Meta may be used to provide you with accurate responses, so don’t send messages to Meta with information you don’t want it to know. '
@alextecplayz The thing I'm talking about is this:
https://engineering.fb.com/2025/04/29/security/whatsapp-private-processing-ai-tools/
The link didn't came across when I translated the Dutch article, I've added it back in. You'll find that this information is newer than the whitepaper you mention (and the whitepaper has no reference to it yet).
As @Wen already pointed out, the FAQ's actually only prove that Meta can grant themselves access to encrypted messages, which was the actual point I was making.
Currently they do so at the request of a participants of a conversation only (according to those FAQ 's). But like "we'll never put ads in WhatsApp", there's nothing stopping them from changing that later.
Considering their 180 degree turn on ads, the fact that they've already controversially started using IG and FB data to train their AI without opt in, what's to stop them from WA being the next step once the infrastructure is there?
I've stopped giving them the benefit of the doubt long ago and they've done nothing to prove me wrong since.
As for the phone number incident, I've never suggested anything about it, other then that it was the thing that got the discovery going. Truth be told I think we'll never know for sure where the AI got it from.
Cc: @helma
You can't even trust them with the content of your messages anymore, because their "AI" will now read them ("privately", of course)
@Wen WhatsApp can read your messages.
On top of that, AI "assistants" from Google and Apple want to scan your messaging apps.
@Wen It is easier than that. They say that but... Can anyone check the code? Nope
So, that's only marketing. And it makes me wonder why they invest money trying to convince us their app is safe. If they can't get any info from the app, why would they be even interested in us using it? 🤔
@Wen I love that their ads keep using a mosaic view of images to say that they can't see anything. The mosaic view is seeing something anyway, so it's already a fail even at trying to say what it intends to say.
At least the text is implied to be encrypted, though it is a reminder that they do see that and there are no guarantees they won't implement backdoors at some stage. (We have only their word that they don't do this...)
Some of the data Meta collects about its users and their contacts is pretty scary stuff. I seriously am unable to understand why so many people are comfortable with it even when you try to tell them. It's especially upsetting when you're one of their contacts.
Johndgeek 🌎
Wen
Gauff 🇪🇺
jtb
Okuna
Pam C
Adam Dalliance
Doerk
Roelant
AlexTECPlayz
stealthradek
Draken BlackKnight
Gabriel Adrian Samfira
oxidand
Jaroslav Svoboda
SirWumpus 👾🍁
Dim Simple
Riquiñez 
Kiru
sheislaurence
CyberFrog
Kurt
Nazo