Hash collision in Apple NeuralHash model

Re:Hash collision in Apple NeuralHash model by UseStrict on Wednesday August 18, 2021 @09:56AM (#28220965)

The _initial_ implementation of client-side scanning has the same initial result, but vastly different futures. Up until now it could be (foolishly or otherwise) assumed that Apple had your personal privacy in mind. Now they have demonstrated a willingness to compromise your local device privacy. Right now it's iCloud only, but I'm sure it's a boolean configuration away somewhere to make it scan everything, regardless of whether it's marked for iCloud upload or not. And I don't think Apple is strong enough to withstand a gagged demand to flip that boolean.

On their cloud services (Apple, Google, MS, etc) they can only get to content that a user has implicitly agreed to share (whether or not they understand the ramifications is another conversation). On your device, there's no more separation.

[ Parent | on HN ]

Re:Hash collision in Apple NeuralHash model by UncleMeat on Wednesday August 18, 2021 @10:33AM (#28221481)

But the discussion here is about malicious actors fraudulently inserting files that trigger alarms. If they have control over your device, "content that a user has implicitly agreed to share" is not a meaningful category.

[ Parent | on HN ]

Re:Hash collision in Apple NeuralHash model by cyanite on Wednesday August 18, 2021 @04:23PM (#28226597)

How have they demonstrated that they will compromise your privacy? They compromise it less than the systems scanning on the cloud, since like this most of the match is only known to the device and not the cloud.

If you don’t trust Apple to not lie, then the entire discussion becomes a bit moot, I think, since then they could pretty much do anything, with or without this system.

[ Parent | on HN ]