[naikrovek]

I seem to recall a recent copyright office decision [0] where it was decided that an AI does not own copyright of its own work, because the output is not the effort of an entity which used intellectual effort to create the output. only a human can own copyright, according to that (US-only) copyright office decision.

this means the output of an AI isn't even considered "work" in the eyes of copyright law, if I am understanding. if the output of copilot is not a "work" then the output of copilot cannot be a "derivative work," and cannot violate copyright.

Courts have repeatedly found that only stuff humans create can be copyrighted.

If GitHub Copilot can't produce a work, it can't violate copyright; only human operators can do that.

This makes the recent GitHub announcement about coming Copilot features make much more sense legally: features which show the origin of the suggestion and which allows a user to select which code licenses to use suggestions from. previously these seemed like things to appease critics, but they're tools to help paying copilot users know what code they're actually using. Nice.

IANAL.

anyway, this lawsuit is gonna fail so effin' hard. lol

[0]: https://www.theverge.com/2022/2/21/22944335/us-copyright-off...

[paulryanrogers]

No one is going to take the AI to court. Just like no one is going to sue the Ctrl+V keys on a keyboard. They're going to sue users of Copilot and or human producers of it.

[naikrovek]

the humans at GitHub have created software which by definition cannot create copyrighted works. if the AI cannot create copyrighted works, it cannot violate copyright.

Humans can create copyrighted works, and therefore can violate copyright.

Given these things, I fail to see how anyone at GitHub could get into any trouble, legally.

Thus, the suit against GitHub/Microsoft will fail, which was the point that I apparently failed to communicate clearly enough.

If anyone is liability-adjacent here, it is copilot users, who must always be mindful of what they write anyway.

[paulryanrogers]

Just calling something AI doesn't copyright-wash the output.

If I write an RE to strip license comments and sell it as AI that doesn't mean I'm innocent of facilitating copyright infringement.

Or if I sell a NN to strip watermarks from a movie broadcast it's unlikely a court is going to throw out the case because "welp it's AI, no copyright".

[naikrovek]

go look up the findings from courts on this and come tell me i’m wrong. computers cannot create copyrighted work alone; the human operating and/or programming the computer is who copyright is granted to.

in the case of Copilot, the output is determined by the keystrokes of the user who uses copilot; that user must decide if what they see is to be used, or if some other piece of code should be used instead, or if they need to provide more input or even just write the thing themselves. Copilot aids the user; it does not run automatically or non-interactively, meaning it alone cannot violate copyright.

no computer or computer program has EVER been granted copyright, anyway. ever. justification for those decisions is that copyrightable work necessarily requires intellectual effort to create, which is something only a human can supply, by definition.

[int_19h]

The question is not whether CoPilot output violates copyright (or rather, that is a separate question that is a concern for users of CoPilot). The question is whether CoPilot itself - the model, that is - constitutes a derived work.

[naikrovek]

that will not pass muster to even be argued in court. i’m so tired of arguing with people about this.

go read some court judgements on PACER for a while, you’ll see.