Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

thewire.in

Post reply on HN
Print view View original

11–20 of 25 posts

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#11

Post by throwaway525142 »

square_usual wrote: post #8

At first I thought the bit about seeing user-reported content was a nothingburger, but this is far, far worse: Artificial intelligence initiates a second set of queues – so-called proactive ones – by scanning unencrypted data that WhatsApp collects about its users and comparing it against suspicious account information and messaging patterns [...] as well as terms and images that have previously been deemed abusive.…

> So the WhatsApp AI can arbitrarily decide to unencrypt your chats and send them for review.

Where do you read the "decide to unencrypt your chats" part? I could only find that they have to decide based on the unencrypted parts of a WhatsApp profile plus other unencrypted data.

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#12

Post by bsaul »

i'm still unable to tell if what they've implemented is some kind of automated copy/pasting from the recipient side (which can always be done by screenshotting the phone and sending it via email), or a server-side backdoor that lets one get cleartext copies of the messages.

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#13

Post by schleck8 »

seanieb wrote: post #7

This story is FUD, pointing out that WhatsApp users can forward messages they received to WhatsApp’s abuse team, when reporting abuse… I’m not a fan of FB, but the series of stories on this reporting tool which has been in the product for years is so bad, from journalists and publications that should know better, that I’m almost certain there’s some level of malice to it. Apple muddying the water with its billion dol…

i wonder what the solution here looks like — not being allowed to report spam etc? Them completely ignoring moderation and going down the instagram path, direct message style?

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#14

Post by seanieb »

square_usual wrote: post #8

At first I thought the bit about seeing user-reported content was a nothingburger, but this is far, far worse: Artificial intelligence initiates a second set of queues – so-called proactive ones – by scanning unencrypted data that WhatsApp collects about its users and comparing it against suspicious account information and messaging patterns [...] as well as terms and images that have previously been deemed abusive.…

No they can’t arbitrarily decrypt messages. This process happens AFTER a user has complained and forwarded the unencrypted messages to WhatsApps abuse team.

You’re first instinct is correct this is a BS nothing burger by journalist that should know better. It’s a stain on Propublica’s reputation.

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#15

Post by schleck8 »

bsaul wrote: post #12

i'm still unable to tell if what they've implemented is some kind of automated copy/pasting from the recipient side (which can always be done by screenshotting the phone and sending it via email), or a server-side backdoor that lets one get cleartext copies of the messages.

pretty sure its neither but manual reports by a user and probably account rating based on how many chats a user has and their phonenumber to prevent large campaigns

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#17

Post by seanieb »

bsaul wrote: post #12

i'm still unable to tell if what they've implemented is some kind of automated copy/pasting from the recipient side (which can always be done by screenshotting the phone and sending it via email), or a server-side backdoor that lets one get cleartext copies of the messages.

Great journalism eh?!

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#18

Post by square_usual »

square_usual wrote: post #8

At first I thought the bit about seeing user-reported content was a nothingburger, but this is far, far worse: Artificial intelligence initiates a second set of queues – so-called proactive ones – by scanning unencrypted data that WhatsApp collects about its users and comparing it against suspicious account information and messaging patterns [...] as well as terms and images that have previously been deemed abusive.…

It's disappointing to see so many people here read maybe the first 10% of the article and decide it's nothing, when there's so much damning about WhatsApp's business model in here. WhatsApp revealed complete metdata that led to a whistleblower getting indicted, they harvest reams of personal data about you, and their AI can decide to unencrypt your information if it feels like it. Even the bit about moderating user-reported content isn't entirely "nothing" either, because WhatsApp's messaging clearly says nobody, including WhatsApp, can read your messages, which is incorrect. This is bad and not holding WhatsApp to account for this, when this very community spent weeks raging at Apple, as weird.

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#19

Post by seanieb »

square_usual wrote: post #18
square_usual wrote: post #8

At first I thought the bit about seeing user-reported content was a nothingburger, but this is far, far worse: Artificial intelligence initiates a second set of queues – so-called proactive ones – by scanning unencrypted data that WhatsApp collects about its users and comparing it against suspicious account information and messaging patterns [...] as well as terms and images that have previously been deemed abusive.…

It's disappointing to see so many people here read maybe the first 10% of the article and decide it's nothing, when there's so much damning about WhatsApp's business model in here. WhatsApp revealed complete metdata that led to a whistleblower getting indicted, they harvest reams of personal data about you, and their AI can decide to unencrypt your information if it feels like it. Even the bit about moderating user-r…

They can’t decrypt your messages. You would first need to send them your unencrypted messages before they can read them… please correct your comment. There’s enough misinformation in the world.

Re: Facebook Undermines Privacy Protections for Its 2B WhatsApp Users

#20

Post by square_usual »

throwaway525142 wrote: post #11
square_usual wrote: post #8

At first I thought the bit about seeing user-reported content was a nothingburger, but this is far, far worse: Artificial intelligence initiates a second set of queues – so-called proactive ones – by scanning unencrypted data that WhatsApp collects about its users and comparing it against suspicious account information and messaging patterns [...] as well as terms and images that have previously been deemed abusive.…

> So the WhatsApp AI can arbitrarily decide to unencrypt your chats and send them for review. Where do you read the "decide to unencrypt your chats" part? I could only find that they have to decide based on the unencrypted parts of a WhatsApp profile plus other unencrypted data.

The article clearly mentions that the AI initiates the proactive queue:

    Artificial intelligence initiates a second set of queues [...] by scanning unencrypted data.
Initiates pretty clearly signals that no user reported the users in this queue. Also, the article mentions before this that the workers have to look at two queues of tickets which have unencrypted content for the moderators to judge. I agree that this is somewhat unclear, but I think it's certainly talking about two different things.

E: the ProPublica article clarified that this is incorrect: https://www.propublica.org/article/how-facebook-undermines-p...

Post reply on HN