> Until you can show an example or paper where the same adversarial image generates a specific hash value for two unrelated perceptual hash functions, with one

> Until you can show an example or paper where the same adversarial image generates a specific hash value for two unrelated perceptual hash functions, with one being hidden, it is not right to predict a high likelihood of that first scenario being possible. "There is no paper attacking ROT13 done twice, therefore it must be secure". Usually, it's on the one proposing the protocol to make a case for its security. Doubly so when it's supposed to last a long time, a lot of people are interested in attacking it, and successful attacks can put people in harm's way. You know what, if you think that this is difficult, feel free to pick an existing perceptual hash function H, cough up some money, and we'll announce a modest prize (say $4000) on HN for the first person to have a working collision attack for NeuralHash+H. H will run on a scaled-down thumbnail, and we'll keep the precise identity of the algorithm secret. If the challenge gets any traction, but nobody succeeds within 40 days, I'll pay you $4000 for your effort. If you're right, this should be easy money. (cf SHA1, which lasted 22 years) Heck If Apple claims that this is difficult (afaict they don't, it would be unwise), they might even join in with their own preimage challenge for $$$. It'd be a no-brainer, a simple and cheap way of generating good publicity.