The implication -- and I think it's a valid one -- is that this client-side mechanism will be very quickly co-opted to also alert on non-CSAM material. For example, Winnie the Pooh memes in China.
I think it’s not valid to claim that it will be quickly used for that purpose.
However I absolutely agree that it could be used to detect non-CSAM images if Apple colludes with that use case.
My point is that this is immaterial to what is going on in China. China is already an authoritarian surveillance state. Even without this, the state has access to the iCloud photo servers in China, so who knows what they are doing, with or without Apple’s cooperation.
> Now people are seriously arguing that continuous searching through your entire life without any warrant or justification by opaque algorithms is fine.
Does anyone seriously doubt that Germany will use this mechanism to ban Nazi imagery?
Then from there, it’s not a big leap to talk about controlling far right (or far left) memes in France or the UK.
More insidiously, suppose some politician in a western liberal democracy was caught with underage kids, and there were blackmail photos that leaked. Do you think those hashes wouldn’t instantly make their way onto this ban list?
I’ll let you change the subject, but let’s note that every time someone realizes privacy in China as a concern, it’s just bullshit.
> Does anyone seriously doubt that Germany will use this mechanism to ban Nazi imagery?
Yes.
> Then from there, it’s not a big leap to talk about controlling far right (or far left) memes in France or the UK.
This one is harder for me to argue against. Those countries could order such a mechanism, whether Apple had built this or not. Because those countries have hate speech laws and no constitutional mechanism protecting freedom of speech.
This is a real problem, but banning certain kinds of speech is popular in these societies. It is disturbingly popular in the US too. That is not Apple’s doing.