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

The deceptive PR behind Apple’s “expanded protections for children”

piotr.is

Post reply on HN
Print view View original

11–20 of 595 posts

Re: The deceptive PR behind Apple’s “expanded protections for children”

#11

Post by samename »

gonehome wrote: post #6
querez wrote: post #2

I have a newborn at home, and like every other parent, we take thousands of pictures and videos of our newest family member. We took pictures of the very first baby-bath. So now I have pictures of a naked baby on my phone. Does that mean that pictures of my newborn baby will be uploaded to Apple for further analysis, potentially stored for indefinite time, shared with law enforcement?

It's worth reading this, which is basically the only good reporting I've seen on this topic: https://daringfireball.net/2021/08/apple_child_safety_initia... There are legitimate things to be concerned about, but 99% of internet discussion on this topic is junk.

John Gruber is biased because his brand is closely tied to Apple’s brand. Ben Thompson wrote a better review on the topic: https://stratechery.com/2021/apples-mistake/

There’s also the Op-Ed by Matthew Green and Alex Stamos, cyber security researchers: https://www.nytimes.com/2021/08/11/opinion/apple-iphones-pri...

Re: The deceptive PR behind Apple’s “expanded protections for children”

#12

Post by czzr »

seph-reed wrote: post #9

I really don't see why the scanning would ever be done on the phone instead of on iCloud if it only affects iCloud images. But I do have guesses why.

Only semi-good reason is it would enable E2E encryption in the cloud while still allowing detection of CSAM.

Re: The deceptive PR behind Apple’s “expanded protections for children”

#13

Post by slownews45 »

gonehome wrote: post #6
querez wrote: post #2

I have a newborn at home, and like every other parent, we take thousands of pictures and videos of our newest family member. We took pictures of the very first baby-bath. So now I have pictures of a naked baby on my phone. Does that mean that pictures of my newborn baby will be uploaded to Apple for further analysis, potentially stored for indefinite time, shared with law enforcement?

It's worth reading this, which is basically the only good reporting I've seen on this topic: https://daringfireball.net/2021/08/apple_child_safety_initia... There are legitimate things to be concerned about, but 99% of internet discussion on this topic is junk.

Even HN reporting / article linking / comments have been surprisingly low quality and seem to fulminate and declaim with surprisingly little interesting conversation and tons of super big assertions.

Linked articles and comments have said apple's brand is now destroyed, that apple is committing child porn felonies somehow with this (the logical jumps and twisting to get to these claims are very far from strong plausible interpretation).

How do you scan for CASM in an E2EE system is the basic question Apple seems to be trying to solve for.

I'd be more worried about the encrypted hash DB being unlockable - is it clear this DOES NOT have anything that could be recreated into an image? I'd actually prefer NOT to have E2EE and have apple scan stuff server side, and keep DB there.

Re: The deceptive PR behind Apple’s “expanded protections for children”

#14

Post by echelon »

gonehome wrote: post #6
querez wrote: post #2

I have a newborn at home, and like every other parent, we take thousands of pictures and videos of our newest family member. We took pictures of the very first baby-bath. So now I have pictures of a naked baby on my phone. Does that mean that pictures of my newborn baby will be uploaded to Apple for further analysis, potentially stored for indefinite time, shared with law enforcement?

It's worth reading this, which is basically the only good reporting I've seen on this topic: https://daringfireball.net/2021/08/apple_child_safety_initia... There are legitimate things to be concerned about, but 99% of internet discussion on this topic is junk.

Gruber practically (no, perhaps actually) worships Apple. He'd welcome Big Brother into his house if it came with an Apple logo, and he'd tell us how we were all wrong for distrusting it. He's not the voice to listen to this time, and you should trust him to have your best interests at heart.

People are furious with Apple, and there's no reason to discount the completely legitimate concerns they have. This is a slippery slope into hell.

It's a good thing congress is about to start regulating Apple and Google. Maybe our devices can get back to being devices instead of spy tools, chess moves, and protection rackets.

(read: Our devices are supposed to be property. Property is something we fully own that behaves the way we want. It doesn't spy on us. Property is something we can repair. And it certainly is not a machination to fleece the industry by stuffing us into walled and taxed fiefdoms, taking away our control. Discard anything that doesn't behave like property.)

[edit: I've read Gruber's piece on this. It's wish-washy, kind of like watching a moderate politician dance on the party line. Not the direct condemnation this behavior deserves. Let's not take his wait and see approach with Dracula.]

Re: The deceptive PR behind Apple’s “expanded protections for children”

#15

Post by gonehome »

cwkoss wrote: post #7
slownews45 wrote: post #5

Earlier quoted context omitted.

If you don't choose upload to icloud, no upload to apple at all. If you do choose icloud upload (most do), they were being uploaded already and stored and may be available to law enforcement. If you do upload to icloud, NOW they will be screened for matches with "known" images in a database, and if you have more than a threshold number of hits, you may be reported. This will happen on device. Apple will also scan pho…

Arent the perceptual hashes based on a chunk of the image? I wonder what the false positive rates are for: - A random image against the DB of perceptual hashes - Images of a baby's skin against the DB of perceptual hashes It seems like the second would necessarily have a higher false positive rate: similar compositions (contains baby's skin) would more likely have similar chunks. Is it just a little higher or several…

It's two factors, both the match on an image hash and an unknown threshold of matches at which point the data gets sent up. If the threshold is not met then nothing gets notified (even if there is a match). Arguably this is why this approach is better for privacy. Cloud matches would not be able to have this extra threshold (in addition to this model allowing e2ee on the cloud in the future).

I'd also like to know more about the specifics here, my guess is that threshold value is pretty high (their 'one in a trillion' comment not withstanding). It's probably targeting large CSAM dumps of matches which would not get flagged by different images.

Re: The deceptive PR behind Apple’s “expanded protections for children”

#16

Post by gambiting »

samename wrote: post #8
querez wrote: post #2

I have a newborn at home, and like every other parent, we take thousands of pictures and videos of our newest family member. We took pictures of the very first baby-bath. So now I have pictures of a naked baby on my phone. Does that mean that pictures of my newborn baby will be uploaded to Apple for further analysis, potentially stored for indefinite time, shared with law enforcement?

Unless those pictures are also in the NCMEC database, there won’t be a match.* * As addressed in the comments below, this isn’t entirely true: the hash looks for visually similar picture and there may be false positives.

Absolutely not true. Apple is using a similarity based hash, so if the NCMEC database contains a picture that's similar to one that you have, it could produce a match even if it's not the same. Apple says this isn't an issue, because a person will look at your picture(yes, a random person somewhere will look at the pictures of your newborn) and judge whether they are pictures of child abuse or not. If this unknown person thinks your picture shows child abuse, you will be reported to NCMEC and then what happens is unknown - but likely that it would result in some legal action against you.

Re: The deceptive PR behind Apple’s “expanded protections for children”

#17

Post by roody15 »

czzr wrote: post #12
seph-reed wrote: post #9

I really don't see why the scanning would ever be done on the phone instead of on iCloud if it only affects iCloud images. But I do have guesses why.

Only semi-good reason is it would enable E2E encryption in the cloud while still allowing detection of CSAM.

Except despite this being repeated over and over… Apple has not said anything about E2E

Re: The deceptive PR behind Apple’s “expanded protections for children”

#18

Post by slownews45 »

cwkoss wrote: post #7
slownews45 wrote: post #5

Earlier quoted context omitted.

If you don't choose upload to icloud, no upload to apple at all. If you do choose icloud upload (most do), they were being uploaded already and stored and may be available to law enforcement. If you do upload to icloud, NOW they will be screened for matches with "known" images in a database, and if you have more than a threshold number of hits, you may be reported. This will happen on device. Apple will also scan pho…

Arent the perceptual hashes based on a chunk of the image? I wonder what the false positive rates are for: - A random image against the DB of perceptual hashes - Images of a baby's skin against the DB of perceptual hashes It seems like the second would necessarily have a higher false positive rate: similar compositions (contains baby's skin) would more likely have similar chunks. Is it just a little higher or several…

Absolute - I think this is one of two key questions for me. That is why I put "known" in quotes. It can't be an exact match because it has to handle cropping, rotation, resize etc.

Images then do get a manual review before a report is made which is good and may help provide feedback on alogs being used.

Going to be hard though for apple to set the second factor to high - I'd say 5 maybe? It's hard to say you had matches on potential CASM and ignored them I'd think.

Re: The deceptive PR behind Apple’s “expanded protections for children”

#20

Post by AlexandrB »

seph-reed wrote: post #9

I really don't see why the scanning would ever be done on the phone instead of on iCloud if it only affects iCloud images. But I do have guesses why.

That's the crux of it. Why bother with on-device identification, unless one of:

a. Apple intends to E2E encrypt iCloud data.

b. This is intended to extend to all photos on the device in the future.

I'm hoping it's (a), but it's probably (b). And in either case it sets a bad precedent for other companies to follow.

Edit: This also turns every jailbreak into a possible CSAM detection avoidance mechanism, giving the government plausible cover to treat them as serious, criminal actions. Apple would probably love that.

Post reply on HN