2 ms·
I think people objected to having the flagged hash database and comparison locally. I suppose the split of perceptual hash client side, comparison server-side m
by brookst 4y ago
I think people objected to having the flagged hash database and comparison locally. I suppose the split of perceptual hash client side, comparison server-side might be objectionable on the "it's draining my battery" angle, but I suspect those who prefer cloud-side scanning would generally like hash creation client-side if it were the price of E2EE.
Of course that's a lot less secure and leaks more privacy info, but the objections seem more about the morality of having a device you own doing the scanning, not so much about any privacy concerns.
- ec109685 4y agoThe comparison didn’t happen locally in Apple’s scheme. Along with each photo a encrypted fingerprint was uploaded and all scanning happened sever side.
- brookst 4y agoThat wasn't my read from https://www.apple.com/child-safety/pdf/CSAM_Detection_Technical_Summary.pdf https://www.apple.com/child-safety/pdf/CSAM_Detection_Techni... : Lots to look at there but the key bit is: >If the user image hash matches the entry in the known CSAM hash list, then the NeuralHash of the user image exactly transforms to the blinded hash if it went through the series of transformations done at database setup time. Based on this property, the server will be able to use the cryptographic header (derived from the NeuralHash) and using the server-side secret, can compute the derived encryption key and successfully decrypt the associated payload data So my understanding is that the server won't even know the hash of the image unless the client-side operations determine that it matches one of the hashes in the CSAM database, which is stored on the client. Are you seeing it differently?
- ec109685 4y agoI’m seeing the same. There are a lot of people saying apple’s scheme was somehow worse than Google’s. I was arguing that apples was at least attempting to be privacy preserving.