>Isn't your bank balance in a bank database also "just a number"?

Absolutely not, but also "yes, which means no". In the first case, a bank balance isn't "just" a number, it's a massively regulated and legally backed number with many layers of interlocking entities, both private and multiple layers of government, in charge of maintenance, auditing, insuring, and enforcing. There is no equivalency to cryptocurrency there, as has been regularly touted.

To the second, it could certainly be argued that a bank balance is indeed "just a number" and that's the point, what gives the number its value is all the infrastructure around it not anything intrinsic to the number itself. If someone finds out my bank balance in Account ABC is $42076 that might have privacy implications sure, but knowing that number gives you access to absolutely nothing of meaning. That's a completely different situation to one where independently finding a given number, which note you need not even have any idea who it belongs to, suddenly equates to ability to make use of that number in real world relevant ways by social consensus.

We're talking more the equivalent of Adam guessing a winning lottery ticket, and then hanging onto it hoping the value will go up and he can trade on the ticket or do other things with it while not actually cashing it in because it's so unlikely somebody else will guess the ticket. Maybe because the lotto ticket winners are published on a public ledger, and Adam doesn't want the notoriety, or at least not just yet. Then Bob does independently guess it, immediately turns it in, and now Adam's lotto ticket is worthless. Bob didn't steal anything from Adam. Whether what Bob did is ok or not depends on the rules of the game.

>I understand that the bank's ownership of its computer means that hacking into it could be seen as (for example) a trespass

Holy shit are you for real? COULD be seen? Yes hacking into a bank would absolutely mean felony prosecution on multiple counts if you were caught.

>However, what if you somehow persuaded a bank employee to change someone's balance?

They would be committing multiple felonies and you would be committing criminal conspiracy, inducement and so on depending on jurisdiction, and probably wire fraud and a bunch of other stuff if you do it remotely that are sorta gimmes for prosecutors.

>The bank employee has some kind of authority to do this and the result is once again "just a number".

The bank employee does not have legal authority to do this. Any technical authority they have is only within the auspices of the law, internal compliance controls and practices and on and on.

Anyway without going through your whole post you're doing a whole lot of false equivalency. Breaking into and modifying somebody else's systems is no small point, it's explicitly illegal under the CFA in the US and similar in the rest of the developed world. There's no such thing as legally "copying" money from an end owner perspective, even if internally to the global financial systems when it comes to fiat currencies from the Treasury & Fed or other national equivalents to banks and other governments and so on it gets more complicated. It's all meant to effectively be a digital version of actual old fashioned hard currency. Hence the entire core concept of theft: it applies to zero sum games, where one person getting more cash means another person now has less.

I'd welcome any actual specific laws on the books about cryptocurrency that contemplates what would happen if someone simply guesses a private key with no interaction with anyone else and then uses it on the network. But without that it's hard to see any existing precedent. On the contrary, cryptocurrency people have repeatedly pushed, and built into the core foundations, the notion of code being law, that possession of a private key is all that's needed and the rest is up to the network and you're supposed to be in charge of that (or someone else is on your behalf and that relationship can be subject to contracts).

> Holy shit are you for real? COULD be seen? Yes hacking into a bank would absolutely mean felony prosecution on multiple counts if you were caught.

I meant to refer specifically to the trespass theory (advocated about 25-30 years ago by some companies as a way to enforce terms of service) as a reason one might attempt to distinguish "changing a number on company X's computer" from "changing a number in a distributed database". That is, there might be legal theories that are more protective of individual companies' computers just because the physical computers belong to the companies as opposed to information-in-general.

https://en.wikipedia.org/wiki/Trespass_to_chattels#Early_app...

However, other forms of computer crime law can protect information-in-general, regardless of where it's stored or by whom.

My point was that existing laws have been happy to punish changing numbers on computers based on the meanings that those numbers have to people, what people act as though those numbers represent. I believe some of these laws are drafted broadly enough that they already treat stealing cryptocurrency as illegal. Even if legislators didn't consciously regulate it this way, courts may conclude that concepts of fraud, property, conversion, etc., already apply to cryptocurrency systems, even if there isn't an obvious technical difference between a transfer intentionally authorized by a human owner and a transfer authorized as a result of fraud, hacking, bugs, etc.

I understand that in, say, Bitcoin, "ownership" of assets stored in a UTXO is implemented only as the ability to cause a transaction that consumes that UTXO, and that this ability doesn't refer to a person's name or identity, or to good or evil, or to the reason that someone caused such a transaction, or to how someone came to possess the necessary information to create it. The blockchain consensus is updated based on whether the transaction followed certain deterministic rules, and concepts like "the owner" do not in fact appear directly anywhere in those rules. However, this doesn't stop a court from saying that some such transactions represent fraud or conversion or something while others don't, even though the transactions in question were equally valid according to the blockchain consensus.

I understand that there's uncertainty and debate in the cryptocurrency world about how we should want legal systems to regulate or not regulate cryptocurrency, remedy or not remedy otherwise-wrongful actions committed via cryptocurrency systems, and enforce or not enforce agreements implemented in or through cryptocurrencies. I also think you're right to point out that there's an issue about whether the content or behavior of the code is, or is meant to be, the "entire agreement" among parties using it, or whether it just somehow reflects other kinds of relationships that are also partly enforced by legal systems.

I currently work on smart contracts for a living. I find the question of how legal systems should view them fascinating, and I don't have a clearly articulated position on it.

Edit: I'd again like to point to Kremen v. Cohen as an analogy. In that case there was a privately (sort of) created database of domain name registrations. There weren't specific laws or regulations created to describe how the courts should view domain name ownership. The defendant in that case fraudulently caused a domain name to be transferred from the plaintiff to the defendant. The courts agreed that the domain name was "property" and that the defendant could be sued for this, again even though there was no specific legislation regulating the domain name industry. Now, many people are unhappy about various ways that the legal systems of various countries try to control and regulate domain name ownership and transfer. I know people who've worked on naming systems that are explicitly meant to be harder for governments to regulate.

Still, when courts looked at the original DNS decades ago, none of these forms of queasiness about the government's role stopped the courts from concluding that domain names were property based on their characteristics and use, and that people could be sued for fraudulently taking domain names away from other people.

It seems like you might be perceiving a kind of hypocrisy in the notion of people wanting to deliberately create things that are harder to regulate, and then still sometimes involving the courts in disputes over them.