Earlier quoted context omitted.
"Apple said that there is a one in one trillion chance of a false positive" https://techcrunch.com/2021/08/05/apple-icloud-photos-scanni... Versus: https://blog.roboflow.com/nerualhash-collision/ Along with: https://github.com/anishathalye/neural-hash-collider Demonstrates blatant incompetence on the part of Apple's design. At this point, trust is broken. That's the end of credibility regarding any other claim made a…
Apple was talking about the chance of accidental false positives. There is a 1 in 4 billion chance that if I send a message to an IP address, it will be yours. The fact that I can know your IP address and send you a message intentionally does not make the previous statement false. > As Apple's hash algorithm is entirely broken and useless, a malicious actor can easily craft a NSFW, but otherwise legal image, that has…
This is a huge difference as not only does Apple facilitate a gain-of-function in terms of potential detriment to the victim of an attack, the newly-enabled vector also indemnifies the attacker.
The effort required to come up with a hash collision in SHA256 is something on the order of the power output of a trillion suns for a billion years (or some other meaninglessly-prohibitively-large amount). Despite such metrics and off-the-shelf solutions being readily attainable within this field Apple has implemented a system that has resulted in a collision within days, followed by a free automated collision generation tool.
This demonstrates absolute gross incompetence in design, and permanently eliminates any trust any sensible individual can ever have in the system despite your protestations.
There is simply nothing you can say to convince any sane person otherwise, the situation has been factually described and it is what it is.