We need a new license that forbids all training. That is the only way to stop big corporations from doing this.
How is that enforceable against the fly-by-night startups?
The current state of the theory that GPL propagates to AI models
131–140 of 314 posts
Re: The current state of the theory that GPL propagates to AI models
#132Genuine question: if I train my model with copyleft material, how do you prove I did? Like if there is no way to trace it back to the original material, does it make sense to regulate it? Not that I like the idea, just wondering. I have been thinking for a while that LLMs are copyright-laundering machines, and I am not sure if there is anything we can do about it other than accepting that it fundamentally changes wha…
> Genuine question: if I train my model with copyleft material, how do you prove I did?
It may produce it when asked
https://chatgpt.com/share/678e3306-c188-8002-a26c-ac1f32fee4...
Re: The current state of the theory that GPL propagates to AI models
#133I honestly think that the most extreme take that " any output of an LLM falls under all the copyright of all its training data" is not really defensible, especially when contrasted with human learning, and would be curious to hear conflicting opinions. My view is that copyright in general is a pretty abstract and artificial concept; thus corresponding regulation needs to justifiy itself by being useful , i.e. encoura…
Human learning is materially different from LLM training. They're similar in that both involve providing input to a system that can, afterwards, produce output sharing certain statistical regularities with the input, including rote recital in some cases – but the similarities end there.
Re: The current state of the theory that GPL propagates to AI models
#134I might be crazy, and I'd love to hear from somebody who knows about this, but I've been assuming that AI companies have been pulling GPL code out of the training material specifically to avoid this. Corporations have always talked about the virality of GPL, sometimes but not always to the point of exaggeration, you'd think that after getting the proof of concept done the AI companies would be running away at full sp…
> I might be crazy, and I'd love to hear from somebody who knows about this, but I've been assuming that AI companies have been pulling GPL code out of the training material specifically to avoid this. Haha no. https://windsurf.com/blog/copilot-trains-on-gpl-codeium-does... And just in the last two days, AI generating LGPL headers (which it could not do if identifying LGPL code was pulled from the codebase) and misat…
Thanks for the links.
That first link shows people actively pulling out GPL code in 2023 and marketing around that fact, though. That's not great evidence that they're not doing it now, especially if testing for if GPL code is still in there seems to be as easy as prompting with an incomplete piece of it.
I'd think that companies could amass a collection of all known GPL code and test for it regularly in order to refine their methods for keeping it out.
> (which it could not do if identifying LGPL code was pulled from the codebase)
Are you sure about this? Linking to LGPL code is fine afaik. And why not train on code that linked to universally available libraries that are legal to use? Seems like one might even prefer it.
Seems like this was rejected for size and slop reasons, not licensing. If the submitter of the PR isn't even fixing possibly hallucinated author's names, it's obvious that they didn't really read it. Debugging vibe coded stuff is like finding an indeterminate number of needles in a haystack.
Re: The current state of the theory that GPL propagates to AI models
#135Earlier quoted context omitted.
I know it's not popular on HN to have anything but supportive statements around GPL, and I'm a big GPL supporter myself, but there is nuance in what is being said here. That case was important, but it's not abojt the virality. There have been no concluded court cases involving the virality portion causing the rest of the code to also be GPL'd, but there are plenty involving enforcement of GPL on the GPL code itself.…
There is no "virality", and the article's use of "propagation" to mean the same thing is wrong. The GPL doesn't "cause" anything to be GPLed that hasn't been explicitly licensed under the GPL by the owner of its copyright. The GPL grants a license to use the copyright material to which it applies. To satisfy the terms of that license for a particular use may require that you license other code under the GPL, but if y…
Virality is a very good feature of GPL and part of what makes it a meaningfully different choice than other open licenses, I don't know why you want attribute that to Microsoft of all places.
Re: The current state of the theory that GPL propagates to AI models
#136Earlier quoted context omitted.
> by the time it's settled Why would the GPL settle? Even more, who is authorized to settle for every author who used the GPL? If the courts decided in favor of the GPL, which I think would be likely just because of the age and pervasiveness of the GPL, they'd actually have to lobby Congress to write an exception to copyright rules for AI. A large part of the infrastructure of the world is built on the GPL, and the p…
I think they meant "settled" as in "resolved."
I meant the same. I don't actually think that the GPL is an entity that can settle a court case; if I meant that I would have said the FSF or something. I mean that in order for it to resolve, a judge has to say that the GPL does not apply.
If ultimately copyright holds up against the models*, the GPL will be a permanent holdout against any intellectual property-wide cross-licensing scheme. There's nobody to negotiate with other than the license itself, and it's not going to say anything it hasn't said before.
* It hasn't done well so far, but Obama didn't appoint any SCOTUS judges so maybe the public has a chance against the corporations there.
Re: The current state of the theory that GPL propagates to AI models
#137Earlier quoted context omitted.
Honest question: why don’t you think it is fair use? I can see how it pushes the boundary, but I can’t lay out logic that it’s not. The code has been publish for the public to see. I’m always allowed to read it, remember it, tell my friends about it. Certainly, this is what the author hoped I would do. Otherwise, wouldn’t they have kept it to themselves? These agents are just doing a more sophisticated, faster versio…
Some project like Wine forbids you to contribute if you ever have seen the source of MS Windows [1]. The meatball inside your head is tainted. I don't remember the exact case now, but someone was cloning a program (Lotus123 -> Quatro or Excel???). They printed every single screen and made a team write a full specification in English. Later another separate team look at the screenshots and text and reimplement it. App…
> I don't remember the exact case now, but someone was cloning a program (Lotus123 -> Quatro or Excel???). They printed every single screen and made a team write a full specification in English. Later another separate team look at the screenshots and text and reimplement it. Apparently meatballs can get tainted, but the plain English text loophole was safe enough.
This is close to how I would actually recommend reimplementing a legacy system (owned by the re-implementer) with AI SWE. Not to avoid copyright, but to get the AI to build up everything it needs to maintain the system over a long period of time. The separate team is just a new AI instance whose context doesn’t contain the legacy the code (because that would pollute the new result). The amplify isn’t too apt though since there is a difference between having something in your context (which you can control and is very targeted) and the code that the model was trained on (which all AI instance will share unless you use different models, and anyways, it isn’t supposed to be targeted).
Re: The current state of the theory that GPL propagates to AI models
#138Earlier quoted context omitted.
There is no "virality", and the article's use of "propagation" to mean the same thing is wrong. The GPL doesn't "cause" anything to be GPLed that hasn't been explicitly licensed under the GPL by the owner of its copyright. The GPL grants a license to use the copyright material to which it applies. To satisfy the terms of that license for a particular use may require that you license other code under the GPL, but if y…
Virality is a very good feature of GPL and part of what makes it a meaningfully different choice than other open licenses, I don't know why you want attribute that to Microsoft of all places.
A key pillar of Microsoft's FUD campaign against open source was that if you use GPL software you run the risk of inadvertantly including some of it in your proprietary software and accidentally causing the whole thing to suddenly become open source against your horrified company's wishes. It was a lie then and it's a lie now. The comment I was replying to (along with many others on this post) indicates the brainworm lives on.
The difference between a license and a contract may be too subtle for the denizens of HN to grasp in 2025 but I assure you it's not lost on the legal system. It's not lost on those of us who followed groklaw back in the day, either. Sad we have to live with an internet devoid of such joys now.
Re: The current state of the theory that GPL propagates to AI models
#139Earlier quoted context omitted.
It's not Microsoft FUD, you're describing the license as viral too, but playing with words. The fact is that if you include GPL'd stuff in your stuff, that assemblage has to conform to the GPL's rules. You're basically saying "the GPL doesn't go back in time and relicense unrelated code." But nobody was ever claiming it does, and describing it as "viral" doesn't imply that it does. It's "viral" because code that you…
It's not just words, except to the extent the law is just words. You said there haven't been any cases involving the "virality portion" but there have. Just not under the "GPL makes other code GPLed" interpretation, because that, as we clearly agree, doesn't exist. What you're calling the "virality portion" says that one of the ways you *are* allowed to use the code is as part of other GPLed software. If you're going…
I do greatly appreciate you talking about cases instead of leaving it at saying there isn't a part of the license and calling any discussion about it FUD.
The Cisco case was about distributing GPL binaries, not linking it with the rest of the code base and the rest of that code base then needing to be GPL. It's a standard license enforcement unrelated to the unique requirements of GPL.
The BusyBox case is probably the closest in the list, but as you already point out we didn't get a ruling to set precedent and instead got a settlement. It seems obvious what the ruling would be (to me at least), but settlement was explicitly not what is being talked about.
Bringing in French courts, they issued fines - they didn't issue the type of order this article talks about which is about releasing the entire thing involved at the time with GPL.
This isn't related to fear, uncertainty, or doubt about GPL. It's related to what has/hasn't already been ruled in the court systems handling this license before as the article skips past a bit. Even in the case we assume the courts will rule with what seems obvious (to me at least), it has a tangible difference in how these cases will be run, the assumptions they will have, and how long they will last.
Re: The current state of the theory that GPL propagates to AI models
#140[flagged]
Copyleft isn't about the software authors freedom, it's about the end-users freedom. Copyleft grants the end-user the freedom to study and modify the code, i.e. the right to repair. Contrast this with closed-source software which may incorporate permissively licensed code: the end-user has no right to study, no right to modify, and no right to repair. Ergo less freedom.