21–30 of 135 posts
The name is hilarious because HME is anything but speedy -- by many orders of magnitude. I think the real fix is secure enclaves, and those have proven to be difficult as well.
This is a massive announcement for AI and use cases related to PII.
This must be the first real world use case of HE. It has generally been considered too slow to do anything useful but this is an excellent use case.
I tried to look homomorphic encryption up casually earlier this year. I saw references that it was being used, but I don’t think they said where. This is one topic I have a very hard time with, I just don’t know enough math to really grok it. It just seems crazy a system could operate on encrypted data (which is effectively random noise from the server’s point of view) and return a result that is correctly calculated…
You have a 7-bit character representation (e.g. ASCII) and your encryption is to add 1 mod 128. E.g. 0 -> 1, 1 -> 2, ... 126 -> 127, 127 -> 0.
As it turns out, all your operations can be represented as adding or subtracting constants. You can now encrypt your data (+1), send it to a remote server, send all the adding and subtracting operations, pull back the processed data, decrypt the data (-1).
Of course, this example is neither useful encryption nor generally useful operation, but can be useful for grokking why it might be possible.
Earlier quoted context omitted.
Edge's Password Monitor feature uses homomorphic encryption to match passwords against a database of leaks without revealing anything about those passwords: https://www.microsoft.com/en-us/research/blog/password-monit... So not the first, but definitely cool to see more adoption!
I believe Safari does the same as well, so not even technically the first at Apple if I’m correct?
I'm sure someone here understands that and can answer conclusively, but that's not me today.
The name is hilarious because HME is anything but speedy -- by many orders of magnitude. I think the real fix is secure enclaves, and those have proven to be difficult as well.
I think Swift in this case is just referring to the programming language, Swift, and not a characteristic of the encryption library itself
This must be the first real world use case of HE. It has generally been considered too slow to do anything useful but this is an excellent use case.
I tried to look homomorphic encryption up casually earlier this year. I saw references that it was being used, but I don’t think they said where. This is one topic I have a very hard time with, I just don’t know enough math to really grok it. It just seems crazy a system could operate on encrypted data (which is effectively random noise from the server’s point of view) and return a result that is correctly calculated…
What we've done here is this:
(a * key) + (b * key) = (c * key)
The rules of elementary algebra allow us to divide out the key on both sides because of a few useful symmetries that addition and multiplication have. Namely, these two equations are always the same number:
(a + b) * key = (a * key) + (b * key)
This is known as the distributive property. Normally, we talk about it applying to numbers being added and multiplied, but there are plenty of other mathematical structures and pairs of operations that do this, too. In the language of abstract algebra, we call any number system and pair of operations that distribute like this a "field", of which addition and multiplication over real[0] numbers is just one of.
A simple example of a field that isn't the normal number system you're used to is a 'finite field'. To visualize these, imagine a number circle instead of a line. We get a finite field by chopping off the number line at some prime[1] number that we decide is the highest in the loop. But this is still a field: addition and multiplication keep distributing.
It turns out cryptography loves using finite fields, so a lot of these identities hold in various cryptosystems. If I encrypt some data with RSA, which is just a pair of finite field exponents, multiplying that encrypted data will multiply the result when I decrypt it later on. In normal crypto, this is an attack we have to defend against, but in homomorphic crypto we want to deliberately design systems that allow manipulation of encrypted data like this in ways we approve of.
[0] Also complex numbers.
[1] Yes, it has to be prime and I'm unable to find a compact explanation as to why, I assume all the symmetries of algebra we're used to stop working if it's not.
The name is hilarious because HME is anything but speedy -- by many orders of magnitude. I think the real fix is secure enclaves, and those have proven to be difficult as well.
> I think the real fix is secure enclaves FTA: “Live Caller ID Lookup uses homomorphic encryption to send an encrypted query to a server that can provide information about a phone number without the server knowing the specific phone number in the request” So, this would require a distributed Secure Enclave or one of them on Apple’s server communicating with one on an Apple device (likely, certainly over time, with lo…
I did actually get that job, but I found out that that interviewer actually said "no", I believe because he thought I was wrong about that.
[1] My usual disclaimer: It's not hard to find my work history, I don't hide it, but I politely ask that you do not post it here directly.
The name is hilarious because HME is anything but speedy -- by many orders of magnitude. I think the real fix is secure enclaves, and those have proven to be difficult as well.
That makes HE anything but Swift (
Earlier quoted context omitted.
> I think the real fix is secure enclaves FTA: “Live Caller ID Lookup uses homomorphic encryption to send an encrypted query to a server that can provide information about a phone number without the server knowing the specific phone number in the request” So, this would require a distributed Secure Enclave or one of them on Apple’s server communicating with one on an Apple device (likely, certainly over time, with lo…
I don't see why it would? IIUC, the promise of homomorphic encryption is that I can encrypt my database of contacts and send it to an untrusted server, later send the encrypted query to that untrusted server, and get back an encrypted response, without the server being able to tell anything that couldn't be told from the wire (some bounds on how much data, timing of communication, that sort of thing) or provide an in…
Homomorphic encryption means you can ask Apple "who is calling me" without Apple knowing who is calling you.