So I recently had a discussion with a copyright academic expert, and based on what he said to me, I'm wondering how they are suing.
Basically, the common conception of all produced work being automatically copyrighted is correct. However, one can only sue for copyright infringement for works that have a registered copyright (one can register after the fact, one is then just limited in the damages they can sue for, i.e. no enhanced damages for violations before registration).
My assumption is that even github copilot is violating peoples copyright, since the copyright for the vast majority of the code is not registered (does anyone actually pay to register the copyright on their code?), it can make it difficult to sue for damages, especially as a form of class action (where presumambly the class that has registered their copyright is minimal in size).
was I informed incorrectly (I'm also wondering, if this is true, how this would impact enforcement of GPL type licenses in the USA, as again, the vast majority of code licensed under it isn't registered, so while one could register it after the fact, how many people really do this?)
It's a bit .. odd. Seems to be partially a complaint, and partially a code review of the couple of examples they cite. Not quite sure what the point is there. Other parts are clearly copy-n-pasted with minor changes from Wikipedia without attribution, which is funny to see in a document basically claiming that's what the other side is doing.
My partner is a paralegal, and just yesterday I was marveling that there's no concept of voting legal briefs, just wholesale copying of blocks of text without attribution. You cite /cases/ but briefs are fair game for what anyone outside the legal profession would call plagiarism.
If you want to bring a suit, you have to register the copyright. You can have someone infringe upon your unregistered copyrighted work and all you have to do to sue them for that is to then register the copyright after the infringement.
Basically, the common conception of all produced work being automatically copyrighted is correct. However, one can only sue for copyright infringement for works that have a registered copyright (one can register after the fact, one is then just limited in the damages they can sue for, i.e. no enhanced damages for violations before registration).
My assumption is that even github copilot is violating peoples copyright, since the copyright for the vast majority of the code is not registered (does anyone actually pay to register the copyright on their code?), it can make it difficult to sue for damages, especially as a form of class action (where presumambly the class that has registered their copyright is minimal in size).
was I informed incorrectly (I'm also wondering, if this is true, how this would impact enforcement of GPL type licenses in the USA, as again, the vast majority of code licensed under it isn't registered, so while one could register it after the fact, how many people really do this?)