mr. h

techconc said:

lkrupp said:

And as AppleInsider and Apple have stated, images are not scanned on the device. But you choose to believe it’s a lie because...?

No, images are scanned on device.  What's funny is that people are just getting worked up about this now.  Apple has been scanning images on our devices for a long time.  This isn't the CSAM hash type scanning, this is the machine learning scanning I'm talking about.  That's how we can search for generic things like "dog" or "beach" a get a bunch of relevant pictures from our library.  Where is all of the "slippery slope" type of discussions around that?  Seriously, the level of stupid being raised about this topic is mind numbing. 

+∞

henrybay said:

Great news! Apple listened. Their CSAM concept made a mockery of Apple’s privacy ethos. Even though it was well intentioned, it would have turned our iPhones into digital Stasi officers monitoring our every move. 

Apple should turn their attention to screening cloud services where much of this offensive material is apparently stored and shared. But they should leave our iPhones alone. Our phones should be sacrosanct paragons of privacy. 

The irony of this post is sky-high.

Their CSAM concept was actually an extremely clever way of enabling all of your photos to be uploaded to iCloud fully encrypted (without giving Apple the keys), such that neither Apple nor anyone else (should they hack into iCloud, or be law-enforcement with a warrant) would have been able to inspect the photos in iCloud, with the exception of any individual photos that matched a CSAM hash, with the proviso that even then, there would have to be at least 30 photos that matched known CSAM material, before even that was possible.

But now, since they have backed down, all of your photos will continue to be uploaded to iCloud unencrypted, where Apple, law enforcement, and any hackers will be able to inspect all of your photos.

Which one of these two scenarios offers the most privacy?

muthuk_vanalingam said:
I don't think your understanding is correct. It is NOT exact copy of of known CSAM that is being searched for/compared against in your phone. There is a level of "pattern matching" involved when CSAM check is done on the phone before it is uploaded to iCloud. So new photos taken will also get flagged IF the pattern matches with known CSAM database entries.

This is not correct. Only known CSAM images, or minor modifications of those known CSAM images, would create a match. It would not be able to detect "new" child porn. Only the kind of AI-based scanning they were proposing for the Messages app on children's phones would theoretically be able to do that (and it's not what Apple was planning to do anyway)

JBSlough said:

Just a question on this tech. So, the software scans for known images of child porn from a data base from a third party. Which would imply that these images are downloaded off the internet or texted to one’s phone. Well, what about the person whose actually using his/hers iPhone to shoot pics, say in their homes? Those wouldn’t be in the database. Exactly how does that work? Or are those images “safe”? It seems to me that this tech only solves half a problem, that is images of known child porn. Not new ones. Am I correct in understanding this? 

Yes, that is correct. The proposed CSAM process would only be able to match to prior-known images. There is some cunningness to try to ensure that even if the image is slightly modified (e.g. cropped, mirrored, rotated slightly) from the known version a match will still be generated, but it absolutely would not trigger on any kind of "new" photo.

just cruisin said:

mr. h said:

just cruisin said:

But anyway, you missed the point.

People hate ads.  People hate pop ups more.  People really, really hate pop ups that know too much about them. 

Apple, don’t make people hate you.

Is it you who's missed the point? The article is concerning a method that allows the user to choose whether or not they receive personalised ads. So what's the problem here? Don't like personalised ads? Turn them off! Simple.

I think it would be wise if you reconsidered attacking others….especially when you are incorrect.  

The “feature” doesn’t allow a user to turn ads off.   It only gives you the choice of receiving ads that are generic in nature or ads that some third party targets against you based on a profile they collect on your identity.  From Apple and the article “Turning on Personalized Ads increases the relevance of ads shown”.

My point still stands.  Many people choose Apple because they protect users from tracking, profiling, etc.  it is a slippery slope to profit on profiling users.  

People hate ads.  People hate pop ups more.  People really, really hate pop ups that know too much about them. 

Apple, don’t make people hate you.

It's rather concerning that you feel that I've attacked you. If you read my post again very carefully, you will see that I asked a question.

The thrust of your point is that people really hate pop-ups that know too much about them. And they can turn the personalisation off. Hence there is no "know too much about them" problem.

Also, I believe that the only thing that is a pop-up, is this one-time thing that lets you select personalised or not. There's nothing here to suggest that any "ads" will be pop-ups.

Lastly, it's unclear if this is going to introduce any more "ads" in Apple's apps. It seems more targeted towards enabling better-targeted app suggestions in the app store.