@Guy Who cares? It can also be copied in the future, where it is public, before traveling into the past. It can only come into existence once, sometime in the past. If it is destroyed the moment it exists, nobody had the time to copy it. In fact if necessary, to make sure one has the original, two time jumps can be made; one immediately after it exists and can be copied, and then another to plant a bomb to destroy it just a millisecond after it is created, along with any machines or humans that might have learned it in that millisecond.

Using time travel to prevent the data from being created is the only way to make the data inaccessible to time travelers: +1. Waiting for it to be created and then making secret copy completely invalidates that solution and also subverts OP's goal: -1.

Encryption doesn’t work if an attacker has access to time travel and therefore literally unlimited time to brute-force the key

I can already imagine a story of competing time travellers in a temporal arms race about deleting and restoring the data. I mean, if you can go back with a physical copy, someone else could go back with their own copy to undelete it.

@11684: Not quite. Unlimited access to time travel does not equate unlimited patience nor practical automation.

@Revolver_Ocelot That's a moot point. Clearly causality is out the window as soon as there is time travel; for example there is no "cause" for me to exist, in my present form, at any time but this moment. Our brain and body changes minutely with every passing moment. The document is just like me: I copy it in 2032, and (just like me and all I learned in my life) it can be transported to 1932, before I was born, and exist with no "cause" to exist. If I cannot do that with a document, I cannot do it with my own self or my own brain or all the things I learned since the destination time.

@Tom: It does not invalidate that solution, it changes history to make the universe evolve without that information, while still allowing the information to be "caused" naturally, if that is a concern. Say in the original history, a new tech becomes quickly widespread and abused; a threat to humanity. The hero needs the tech to exist but with changes so it can't be abused. So it still must be invented, but not released until proper security measures, not even invented at the time the invention was created, can be employed. Our hero wants to keep the discovery private, not destroy it.

The hard part regarding encryption/decryption is using a technique, that is future proof. We already have a few algorithms, that were secure at the time the algorithm was created like md5, which is now unsecure due to our better technologies in the present. Even our current technology in encryption (e.g. sha256) seems to be breakable by Shor's Algorithm with improvements in quantum computing in the future. To be honest, I think that is an interesting premise, too

@M.Winkens Shor's algorithm is just faster, it is not instantaneous, and the solution to that is simple: longer encryption keys. There is no reason encryption keys cannot be arbitrarily long, and ironically Shor's algorithm and quantum computing will also improve the speed of finding much much longer encryption keys. In this case, the same tech that creates a threat will create the defense against that threat.

What's the point of destroying it after creation? If you're going to change history in a way that would prevent you in the future from acquiring something, there doesn't seem to be a functional difference (as far as time travel is concerned) between destroying it after it's created or preventing it from being created in the first place. You're not addressing the paradox in any way by waiting for it to be created, because the paradox is that you in the future wouldn't have acquired the thing, whether it would've existed or not.

@NotThatGuy Two scenarios: No point in waiting to prevent it from ever existing. But if you want to use it, but keep it from being widely known, then the paradox avoided is having the information originate from nowhere. Creation can involve all kinds of random elements that will never occur again, the right people with the right experiences in the right place at the right time. By allowing the information to exist and THEN stealing it and destroying it, you prevent it from being disseminated and learned, without any paradox of "Where did it come from in the first place?"

I feel there's a hook here somewhere with the encryption key being paradoxical in itself such that attempting to decrypt pops your timeline off - a Diffie-Hellman where the second encryption step is performed in the future, or something along those lines. Dedicated services would provide you with time-looped cryptography for this purpose. I can't quite pin it down, though - I guess my brain is too causal for time shenanigans.

If you follow that process then the paper will always have 1 on it, because you did not write 2 until after you tested that 1 did not work, which is after time B. Instead, you would have to travel back to time A (plus a tiny bit) and write x+1 then. Which only works if the time travel is such that you always travel back to a new timeline where the time travel did not occur (or you collide with yourself) -- but in that case the original you will never get the answer, only the one actually doing the time travel. Thus a time travelling computer is not helpful either. The paper is also redundant.

@Miral Obviously if from my future vantage point I know what needs to modified in the information, I can preserve that as well, before hiding the original. What you are talking about is not a concern. If I stop it from being known in 1982, and in 1992 it is known again, I can figure out why it is known in 1992, and go back and fix that. Or if I know it began in 1982, but I see that what began then is not in it's final form, then I can find out when in time between 1982 and 2022 it was "fixed", and hide all that work so it is retrievable before I stop the original from being known.

IMHO, if you were to do what this answer says, the copies you have, and your memories of it existing, would also disappear. So good luck explaining to yourself why you are there after you've done the deed. If this creates further paradox, well, that's perhaps your clue that traveling back in time is utter BS. Physicists will tell you its not possible (it would require working with imaginary numbers), and logic is trying very hard to tell you the same thing. This is why its one of my 3 least favorite SciFi tropes. Basically its an author dividing by 0 to dispense with all rules and let 1=0.

So now the criminal just goes back to prevent the prevention. I guess that could make for some dramatic fight scenes at least, but it doesn't create permanent security.

@Amadeus while this is true it misses my point. My point is you can't make encryption truely future proof. There might be an issue in it's current implementation, too, where we are not yet aware of. Also the tech to create much better encryption keys is not available in the present yet which further proofs my point

@M.Winkens That is not true, I can make encryption keys of any length. I've done it. I can also double or triple encrypt with multiple one-use keys; just encrypt the encrypted text. If you can't break all of them, you can't break the message. And that won't matter: The criminal would have to find my hidden encrypted file. The idea they can also know the information at the same nanosecond as me is a bit too magical for my taste, and once I prevent it from being disseminated, how do they ever know it even exists? I 100% wiped it out of the future timeline, the world evolves without it.

@Amadeus it appears the message I was replying to was deleted or at least didn't survive the transition into chat; I was referring to someone's description of how to try to brute force an encryption key using time travel, but which wouldn't have worked as described. Really, whether any method works depends on the particular rules of your time travel, which you can mostly define however you need to avoid these sorts of issues, as long as you keep them self-consistent.

Some problems are avoided if time travel necessarily requires travel in person to a location and detecting or copying something with human senses -- that means you can physically defend the location and times when the information is available unencrypted (e.g. displayed on a monitor while you're reading it yourself). If remote time viewing is possible (invisibly spying on past events via wormholes or whatever), then all bets are off unless you can somehow construct a barrier to such observation.