Can someone explain to me how this "sets a precedent for future scans"? I've been following this for weeks but still don't really understand this point.
The same applies to all the other examples mentioned - persecution of political activists is one that comes up a lot. If I were such an activist (or insert any other persecuted demographic) and I have photos of me at some demonstration or rally (or insert any other compromising activity), how can a match by flagged without the actual photos first being in the hashed database?
Since political activists may collect and share known images and memes (think of all the Tiananmen Square photos that still make the round), there are certainly valid concerns that the same algorithms could be used to scan for these instead of CSAM, but it's a stretch. These people would also have to be storing those photos in iCloud Photo Library.
Most of the actual privacy advocates and activists are focused on the examples of what
could happen, and for some of them and the work they do, that's totally fair, but it's not something that's going to directly affect most people.
Almost everybody else who is concerned about this either doesn't fully understand what's going on (e.g., "Apple is going to be scanning everything on my iPhone!"), is playing fast and loose with extreme "what if" scenarios (e.g., "What if Apple suddenly changed the algorithm and all the rules to do something completely different?"), or they're simply philosophically opposed to on-device scanning in any form.
Is it the potential to compromise the hash comparison algorithm such that less and less exact matches can be garnered? This to me seems to be the only way to "trick" the system into reporting my unique photos as matching the CSAM (or other) database, however, I don't see this one articulated too clearly in any of the arguments presented. Also I don't see this as being particularly efficient since in order the ensure retrieval of any targeted photo the bar would have to be set so low that all photos would be most likely be reported and sent for investigation.
Due to the way the hashing algorithms work, it wouldn't be possible to get "less exact" matches. Loosening the algorithm would just result in more false positives, which would end up being completely unrelated photos.
Some folks are also conflating this with the Communication Safety in Messages feature, which is fair as Apple announced both features at the same time, mixing up the messaging, but it's a completely separate feature. Communication Safety does use machine learning to identify sexually explicit photos and blur them out for users under 18 years of age, but it doesn't report anything at all back to Apple. At most, parents can get notified if their kids below the age of 13 actually view or send explicit photos. The whole thing is opt-in, however, and can only be used by those in an iCloud Family Sharing group.
That said, some do fear that once Apple has opened the door to this kind of on-device scanning
and reporting, it could very well choose to build algorithms that would scan for a lot more, but again that's getting into extreme "What if?" scenarios. If you're willing to go down that road, you should have stopped using an iPhone years ago, as Apple "could" do just about anything it wants to behind your back.