Relicensing with AI-Assisted Rewrite

AI can't claim ownership, humans can't either as they haven't produced it. If there is guaranteed no one which can claim ownership it often seen as being in the public domain.

In general it is irrelevant what the copyright of the AI training data is. At least in the US judges have been relevant clear about that. (Except if the AI reproduced input data close to verbatim. _But in general we aren't speaking about AI being trained on a code base but an AI using/rewriting it_.)

(1): Which isn't the same as no one seems to know who has ownership. It also might be owned by no-one in the sense that no one can grant you can copyright permission (so opposite of public domain), but also no-one can sue (so de-facto public domain).

There is no such thing as the output of an LLM as a 'new' work for copyright purposes, if it were then it would be copyrightable and it is not. The term of art is 'original work' instead of 'new'.

The bigger issue will be using tools such as these and then humans passing off the results as their own because they believe that their contribution to the process whitewashes the AI contributions to the point that they rise to the status of original works. "The AI only did little bits" is not a very strong defense though.

If you really want to own the work-product simply don't use AI during the creation. You can use it for reviews, but even then you simply do not copy-and-paste from the AI window to the text you are creating (whether code or ordinary prose isn't really a difference).

I've seen a copyright case hinge on 10 lines of unique code that were enough of a fingerprint to clinch the 'derived work' assessment. Prize quote by the defendant: "We stole it, but not from them".

There is a very blurry line somewhere in the contents of any large LLM: would a model be able to spit out the code that it did if it did not have access to similar samples and to what degree does that output rely on one or more key examples without which it would not be able to solve the problem you've tasked it with?

The lower boundary would be the most minimal training set required to do the job, and then to analyze what the key corresponding bits were from the inputs that cause the output to be non-functional if they were dropped from the training set.

The upper boundary would be where completely non-related works and general information rather than other parties copyrighted works would be sufficient to do the creation.

The easiest way to loophole this is to copyright the prompt, not the work product of the AI, after all you should at least be able to write the prompt. Then others can re-create it too, but that's usually not the case with these AI products, they're made to be exact copies of something that already exists and the prompt will usually reflect that.

That's why I'm a big fan of mandatory disclosure of whether or not AI was used in the production of some piece of text, for one it helps to establish whether or not you should trust it, who is responsible for it and whether the person publishing it has the right to claim authorship.

Using AI as a 'copyright laundromat' is not going to end up well.

it depends on the country you are in

but overall in the US judges have mostly consistently ruled it as legal

and this is extremely unlikely to change/be effectively interpreted different

but where things are more complex is:

- model containing training data (instead of generic abstractions based on it), determined by weather or not it can be convinced to produce close to verbatim output of the training data the discussion is about

- model producing close to verbatim training data

the later seems to be mostly? always? be seen as copyright violation, with the issue that the person who does the violation (i.e. uses the produced output) might not known

the former could mean that not just the output but the model itself can count as a form of database containing copyright violating content. In which case they model provider has to remove it, which is technically impossible(1)... The pain point with that approach is that it will likely kill public models, while privately kept models will for every case put in a filter and _claim_ to have removed it and likely will get away with it. So while IMHO it should be a violation conceptually, it probably is better if it isn't.

But also the case the original article refers to is more about models interacting/using with code base then them being trained on.

(1): For LLMs, it is very much removable for knowledge based used by LLMs.

The main analogy is this one: you take a massive pile of copyrighted works, cut them up into small sections and toss the whole thing in a centrifuge, then, when prompted to produce a work you use a statistical method to pull pieces of those copyrighted works out of the centrifuge. Sometimes you may find that you are pulling pieces out of the laundromat in the order in which they went in, which after a certain number of tokens becomes a copyright violation.

This suggests there are some obvious ways in which AI companies can protect themselves from claims of infringement but as far as I'm aware not a single one has protections in place to ensure that they do not materially reproduce any fraction of the input texts other than that they recognize prompts asking it to do so.

So it won't produce the lyrics of 'Let it be'. But they'll be happy to write you mountains of prose that strongly resembles some of the inputs.

The fact that they are not doing that tells you all you really need to know: they know that everything that their bots spit out is technically derived from copyrighted works. They also have armies of lawyers and technical arguments to claim the opposite.

but also probably not fully right

as far as I understand they avoid the decision of weather an AI can produce creative work by saying that the neither the AI nor it's owner/operator can claim ownership of copyright (which makes it de-facto public domain)

this wouldn't change anything wrt. derived work still having the original authors copyright

but it could change things wrt. parts in the derived work which by themself are not derived