Child abuse is a terrible problem, but big tech and politicians are using it as leverage to erode all of our right to privacy. What's next, let's monitor all text communication to stop perverts texting with your child? It will be hailed as another great step towards fighting paedophilia.
The problem with the "child abuse" angle is that protesting against these changes turns everybody against you, because normal people don't have anything to hide. "Why do you hate the children?"
We are losing all of our right to privacy with a massive applause.
Once people internalise the "nothing to hide" argument they begin to reach for the question "what do you have to hide that makes you think this isn't ok" before they are even prompted with the usual arguments about it only being for finding drugs, guns, terrorists, pedophiles, etc
The fact this is on device and off by default is good because that's how this sort of customer feature should be built, like being able to install a DNS filter or other website filtering software. If you're worried, these options should be available to protect your children... but sadly its a small step from "these are available" to "why didn't the government make them turn it on before I bought it for my child and didn't turn it on before giving it to my child"
There is now a feature that scans every photo you receive. There previously was not something scanning every photo you receive. What used to be a conversation between 2 people is now a conversation between 2 people and an Apple blackbox (assuming you turn the feature on).
As with most of these types of tech, the proponents of the "this involves erosion of privacy" aren't necessarily concerned with the exact implementation as described on release, but with how the blackbox works and how the blackbox (and the laws/regulations/obligations around it) will change in the future, and the inability to change the settings of the blackbox, effectively letting Apple become the touchstone for what is 'appropriate', even if it doesn't align with your view of 'appropriate'.
1) Seen the automatic-text-recognition-on-images features in recent iOS? It's already scanning ~every photo you look at, and not just in iMessage. If you put an image in Photos, it's also doing some object recognition on it (which you can see if you search for e.g. "books" or "chairs" in Photos). Dunno if it does that elsewhere, but the text-recognition thing, yeah, that happens all over the OS in all kinds of apps. Maybe (probably) you can turn that off, but this feature folks are complaining about is opt-in so that shouldn't be relevant for this line of reasoning.
And furthermore:
> Messages uses on-device machine learning to analyze image attachments and determine if a photo appears to contain nudity. The feature is designed so that Apple doesn’t get access to the photos.
2) If you don't like this because you don't trust this statement to be true, then I have some bad news for you about who writes the entire closed-source operating system that runs on iOS devices and are the only ones who can deploy any code they like to them, basically at any time.
This is an opt in feature so it’s up to parents whether they think it’s appropriate or not, but I think it’s weird that people are getting mad at Apple for providing a secure, private way to do this when child abuse online is a real thing that happens and some parents may believe that it’s appropriate to trade off some of their young children’s privacy in exchange for some small degree of protection against that.