Copyright infringement is not theft. These are two completely different issues. When data is copied it is not taken away from the owner like when physical goods are stolen. Secondary damages may or may not occur, but they are not the same as depriving someone of a good. As an analogy, I wouldn't steal a car, but I surely would copy a car if I could do so by simply pressing a button...
This isn't necessarily only about copyright infringement (though it's definitely that too). If some of the source code on your machine contain sensitive information, like API keys, database passwords, etc.
Legally, the word "theft" isn't only used when one party loses anything; a victim of identity theft doesn't lose their identity, yet we don't call it "identity infringement". I'm not familiar enough with US law to know for sure, but it wouldn't surprise me if the word "theft" is used somewhere for obtaining sensitive information without permission.
That still leaves corporate espionage, which (last I checked) is a very severe offense. If that "source code" contained significantly-sensitive data (like medical info or info about legal cases), then there's a giant can of worms right there (and each of those worms has a surname of "Felony").
Copyright Infringement is an act, and at least here in the US, an act which both criminal and civil laws provides specific penalties/remedies. On the criminal side, obviously, one of the penalties is imprisonment.
Ah, I was unfamiliar with criminal penalties for copyright infringement. Could you go ahead and link me to the relevant US Code text that provides for such penalties?
But that's about as unlikely as the code containing trade secrets.
Plus:
- For copyright infringement, they'd need to actually redistribute the code. Using it for machine learning and distributing short snippets wouldn't be copyright infringement.
- For that trade secret stuff you'd need to prove intent.
For copyright infringement, they'd need to actually redistribute the code.
IANAL, but I don't think so. In MAI v. Peak[1], the court determined that even loading a program from disk to RAM was a copy, and therefore infringing without a license. Congress has since then added a specific exception for "Machine maintenance and repair", but that's it. Copying from a remote machine and storing it in their disks should certainly qualify.
> But that's about as unlikely as the code containing trade secrets.
Unpublished code, is itself a trade secret. Even just the processes, procedures, organisation, tooling, library use, etc in the code provides a competitive advantage. i.e. The 'metadata' is also a trade secret.
The only intent you'd need to prove is that the accused is using the trade secret to the 'economic benefit of anyone other than the owner'.
It seems obvious that Kite is training a proprietary ML algorithm, with trade secrets, for their own economic benefit.
Maybe even corporate espionage.