I believe it is a narrow view of the situation. If we take a look into the history, into the reasons for inventing GPL, we'll see that it was an attempt to fight copyrights with copyrights. The very name 'copyleft' is trying to convey the idea.
What AI are eroding is copyright. You can re-implement not just a GPL program, but to reverse engineer and re-implement a closed source program too, people have demonstrated it already, there were stories here on HN about it.
AI is eroding copyright, so there may no longer be a need for the GPL. GNU should stop and rethink its stance, chuck away the GPL as the main tool to fight evil software corporations and embrace LLM as the main weapon.
LLM's - to date - seem to require massive capital expenditures to have the highest quality ones, which is a monumental shift in power towards mega corporations and away from the world of open source where you could do innovative work on your own computer running Linux or FreeBSD or some other open OS.
I don't think that's an exciting idea for the Free Software Foundation.
Perhaps with time we'll be able to run local ones that are 'good enough', but we're not there yet.
There's also an ethical/moral question that these things have been trained on millions of hours of people's volunteer work and the benefits of that are going to accrue to the mega corporations.
Edit: I guess the conclusion I come to is that LLM's are good for 'getting things done', but the context in which they are operating is one where the balance of power is heavily tilted towards capital, and open source is perhaps less interesting to participate in if the machines are just going to slurp it up and people don't have to respect the license or even acknowledge your work.
> LLM's - to date - seem to require massive capital expenditures to have the highest quality ones, which is a monumental shift in power towards mega corporations and away from the world of open source
Yeah, a bit of a conundrum. But I don't think that fighting for copyright now can bring any benefits for FOSS. GNU should bring Stallman back and see whether he can come with any new ideas and a new strategy. Alternatively they could try without Stallman. But the point is: they should stop and think again. Maybe they will find a way forward, maybe they won't but it means that either they could continue their fight for a freedom meaningfully, or they could just stop fighting and find some other things to do. Both options are better then fighting for copyright.
> There's also an ethical/moral question that these things have been trained on millions of hours of people's volunteer work and the benefits of that are going to accrue to the mega corporations.
I want a clarify this statement a bit. The thing with LLM relying on work of others are not against GPU philosophy as I understand it: algorithms have to be free. Nothing wrong with training LLMs on them or on programs implementing them. Nothing wrong with using these LLMs to write new (free) programs. What is wrong are corporations reaping all the benefits now and locking down new algorithms later.
I think it is important, because copyright is deemed to be an ethical thing by many (I think for most people it is just a deduction: abiding the law is ethical, therefore copyright is ethical), but not for GNU.
Is massive capital expenditure not also required to enforce the GPL? If some company steals your GPLed code and doesn't follow the license, you will have to sue them and somebody will have to pay the lawyers.
> Is massive capital expenditure not also required to enforce the GPL?
It's nowhere near the order of magnitude of the kind of spending they're sinking into LLM's. The FSF and other groups were reasonably successful at enforcing the GPL, operating on a budget 1000's of times smaller than that of AI companies.
> LLM's - to date - seem to require massive capital expenditures to have the highest quality ones
There are near-SOTA LLM's available under permissive licenses. Even running them doesn't require prohibitive expenses on hardware unless you insist on realtime use.
> There's also an ethical/moral question that these things have been trained on millions of hours of people's volunteer work and the benefits of that are going to accrue to the mega corporations.
This was already the case and it just got worse, not better.
At a certain point, I think we had reached a kind of equilibrium where some corporations were decent open source citizens. They understood that they could open source things like infrastructure or libraries and keep their 'crown jewels' closed. And while Stallman types might not have been happy with that, it seemed to work out for people.
Now they've just hoovered up all the free stuff into machines that can mix it up enough to spit it out in a way that doesn't even require attribution, and you have to pay to use their machine.
AI essentially gatekeeps all of open source to companies to pluck from to their hearts content. And individual contributors using these tools and freely mixing it with their own - usual minor - contributions are another step of whitewashing because they're definitely not going to own up to writing only 5% of the stuff they got paid for.
Before we had RedHat and Ubuntu, who at least were contributing back, now we have Microsoft, Anthropic and OpenAI who are racing to lock the barn door around their new captive sheep. It's just a massive IP laundromat.
Copyleft is a mirror of copyright, not a way to fight copyright. It grants rights to the consumer where copyright grants rights to the creator. Importantly, it gives the end-user the right to modify the software running on their devices.
Unfortunately, there are cases where you simply can't just "re-implement" something. E.g., because doing so requires access to restricted tools, keys, or proprietary specifications.
"So, I looked for a way to stop that from happening. The method I came up with is called “copyleft.” It's called copyleft because it's sort of like taking copyright and flipping it over. [Laughter] Legally, copyleft works based on copyright. We use the existing copyright law, but we use it to achieve a very different goal."
"very different goal" isn't the same as "fundamentally destroying copyright"
the very different goal include to protect public code to stay public, be properly attributed, prevent companies from just "sizing" , motivate other to make their code public too etc.
and even if his goals where not like that, it wouldn't make a difference as this is what many people try to archive with using such licenses
this kind of AI usage is very much not in line with this goals,
and in general way cheaper to do software cloning isn't sufficient to fix many of the issues the FOSS movement tried to fix, especially not when looking at the current ecosystem most people are interacting with (i.e. Phones)
---
("sizing"): As in the typical MS embrace, extend and extinguish strategy of first embracing the code then giving it proprietary but available extensions/changes/bug fixes/security patches to then make them no longer available if you don't pay them/play by their rules.
---
Through in the end using AI as a "fancy complicated" photocopier for code is as much removing copyright as using a photocopier for code would. It doesn't matter if you use the photocopier blind folded and never looked at the thing you copied.
That’s not a rebuttal of the OP’s point. None of that says anything about fighting copyright. It literally says he flipped it which is wha the OP said when they said it’s a mirror.
> AI is eroding copyright, so there may no longer be a need for the GPL. GNU should stop and rethink its stance, chuck away the GPL as the main tool to fight evil software corporations and embrace LLM as the main weapon.
Is this LLM thing freely available or is it owned and controlled by these companies? Are we going to rent the tools to fight "evil software corporations"?
Its purpose "if you run the software you should be able to inspect and modify that software, and to share those modifications with your peers" not explicitly resist copyright. Yes copyright is bad in that it often prevents one from doing that, but it is not the purpose of the GPL to dismantle copyright.
Reducing it to "well you can clone the proprietary software you're forced to use by LLM" is really missing the soul of the GPL.
> we'll see that it was an attempt to fight copyrights with copyrights
it's not that simple
yes, GPLs origins have the idea of "everyone should be able to use"
but it also is about attribution the original author
and making sure people can't just de-facto "size public goods"
the kind of AI usage is removing attribution and is often sizing public goods in a way far worse then most companies which just ignored the license did
so today there is more need then ever in the last few decades for GPL like licenses
So not only are we moving goalposts here, but we've decided the GNU team should join the other team? I don't understand how GNU would see mass model LLM training as anything but the most flagrant violations of their ethos. LLM labs, in their view, would be among the most evil software corporations to have ever existed.
That's naive. Copyright doesn't just apply to software. There already have been countless lawsuits about copying music long before the term "open source" was invented. No, changing the lyrics a bit doesn't circumvent copyright. Nor does translating a Stephen King novel to German and switching the names of the places and characters.
A court ordered the first Nosferatu movie to be destroyed because it had too many similarities to Dracula. Despite the fact that the movie makes rather large deviations from the original.
If Claude was indeed asked to reimplement the existing codebase, just in Rust and a bit optimized, that could well be a copyright violation. Just like rephrasing A Song ot Ice and Fire a bit, and switching to a different language, doesn't remove its copyright.
Claude was asked to implement a public API, not an entire codebase. The definition of a public API is largely functional; even in an unusually complex case like the Java standard facilities (which are unusually creative even in the structure and organization of the API itself) the reimplementation by Google was found to be fair use.
> Claude was asked to implement a public API, not an entire codebase.
Allegedly. There have been several people who doubted this story. So how to find out who is right? Well, just let Claude compare the sources. Coincidentally, Claude Opus 4.6 doesn't just score 75.6% on SWE-bench Verified but also 90.2% on BigLaw Bench.
It's like our copyright lawyer is conveniently also a developer. And possibly identical to the AI that carried out the rewrite/reimplemention in question in the first place.
While I personally agree with you, Richard Stallman (the creator of the GPL) does not. He has always advocated in favor of strong copyright protection, because the foundation of the GPL is the monopoly power granted by copyright. The problem that the GPL is intended to solve is proprietary software.
Generative models (AI) are not really eroding copyright. They are calling its bluff. The very notion of intellectual property depends on a property line: some arbitrary boundary where the property begins and ends. Generative models blur that line, making it impractical to distinguish which property belongs to whom.
Ironically, these models are made by giant monopolistic corporations whose wealth is quite literally a market valuation (stock price) of their copyrights! If generative models ever become good enough to reimplement CUDA, what value will NVIDIA have left?
The reality is that generative models are nowhere near good enough to actually call the bluff. Copyright is still the winning hand, and that is likely to continue, particularly while IP holders are the primary authors of law.
---
This whole situation is missing the forest for the trees. Intellectual Property is bullshit. A system predicated on monopoly power can only result in consolidated wealth driving the consolidation of power; which is precisely what has happened. The words "starving artist" ring every bit as familiar today as any time in history. Copyright has utterly failed the very goals it was explicitly written with.
It isn't the GPL that needs changing. So long as a system of copyright rules the land, copyleft is the best way to participate. What we really need is a cohesive political movement against monopoly power; one that isn't conveniently ignorant of copyright as its most significant source.
> Blanchard's account is that he never looked at the existing source code directly. He fed only the API and the test suite to Claude and asked it to reimplement the library from scratch
This feels sort of like saying "I just blindly threw paint at that canvas on the wall and it came out in the shape of Mickey Mouse, and so it can't be copyright infringement because it was created without the use of my knowledge of Micky Mouse"
Blanchard is, of course, familiar with the source code, he's been its maintainer for years. The premise is that he prompted Claude to reimplement it, without using his own knowledge of it to direct or steer.
> Blanchard is, of course, familiar with the source code, he's been its maintainer for years.
I would argue it's irrelevant if they looked or didn't look at the code. As well as weather he was or wasn't familiar with it.
What matters is, that they feed to original code into a tool which they setup to make a copy of it. How that tool works doesn't really matter. Neither does it make a difference if you obfuscate that it's an copy.
If I blindfold myself when making copies of books with a book scanner + printer I'm still engaging in copyright infringement.
If AI is a tool, that should hold.
If it isn't "just" a tool, then it did engage in copyright infringement (as it created the new output side by side with the original) in the same way an employee might do so on command of their boss. Which still makes the boss/company liable for copyright infringement and in general just because you weren't the one who created an infringing product doesn't mean you aren't more or less as liable of distributing it, as if you had done so.
So, let's say that rather than actually touching any copyrighted material, a human merely tells an AI about how to go onto the internet and find copyrighted material, download it, and ingest it for training. The AI, fully autonomously, does so, and after training itself on the material deletes it so no human ever downloads, consumes, or shares it.
If we are saying AI is "more than a tool", which seems to be the case courts are leaning since they've ruled AI output without direct human involvement is not copyrightable[0], then the above seems like it would be entirely legal.
Copyright protects even very abstract aspects of human creative expression, not just the specific form in which it is originally expressed. If you translate a book into another language, or turn it into a silent movie, none of the actual text may survive, but the story itself remains covered by the original copyright.
So when you clone the behavior of a program like chardet without referencing the original source code except by executing it to make sure your clone produces exactly the same output, you may still be infringing its copyright if that output reflects creative choices made in the design of chardet that aren't fully determined by the functional purpose of the program.
If you pirate a movie and reencode it, does that apply as well? You can still watch the movie and it is “obviously” the same movie. Here you can use the program and it is, to the user, also the same.
What does derivative mean here? Because IMO it means that the existing work was used as input. So if you used a LLM and it was trained on the existing work, that's a derivative work. If you rot13 encode something as input, so you can't personally read it, and then a device decides to rot13 on it again and output it, that's a derivative work.
In order for it to be creatively derivative you would need to copy the structure, logic, organization, and sequence of operations not just reimplement the functionality. It is pretty clear in this case that wasn't done.
As a cynical person I assume all the frontier LLMs were trained on datasets that include every open source project, but as a thought experiment, if an LLM was trained on a dataset that included every open source project _execept_ chardet, do you think said LLM would still be able to easily implement something very similar?
Of course, the problem with this interpretation is that all modern LLMs are derivatives from huge amounts of text under completely different licenses, including "All rights reserved", and therefore can not be used for any purpose.
I'm not sure how you square the circle of "it's alright to use the LLM to write code, unless the code is a rewrite of an open source project to change its license".
LLMs do not encode nor encrypt their training data. The fact they can recite training data is a defect not a default. You can understand this more simply by calculating the model size as an inverse of a fantasy compression algorithm that is 50% better than SOTA. You'll find you'd still be missing 80-90% of the training data even if it were as much of a stochastic parrot as you may be implying. The outputs of AI are not derivative just because they saw training data including the original library.
Then onto prompting: 'He fed only the API and (his) test suite to Claude'
This is Google v Oracle all over again - are APIs copyrightable?
> This is Google v Oracle all over again - are APIs copyrightable?
Yes this is the best way to ask the question. If I take a public facing API and reimplement everything, whether it's by human or machine, it should be sufficient. After all, that's what Google did, and it's not like their engineers never read a single line of the Java source code. Even in "clean room" implementations, a human might still have remembered or recalled a previous implementation of some function they had encountered before.
I just don't see how it's relevant whether he did look or didn't. In my opinion, it's not just legally valid to make a re-implementation of something if you've seen the code as long as it doesn't copy expressive elements. I think it's also ethically fine as well to use source code as a reference for re-implementing something as long as it doesn't turn into an exact translation.
Right. The alternative is that we reward Dan for his 14 years of volunteer maintenance of a project... by banning him from working on anything similar under a different license for the rest of his life.
It's actually not legally fine, or at least it's extremely dangerous. Projects that re-implement APIs presented by extremely litigious companies specifically do not allow people who, for instance, have seen the proprietary source code to then work on the project.
I don't think fear or legal action makes it illegal.
If I know it is legal to make a turn at a red light. And I know a court will uphold that I was in the right but a police officer will fine me regardless and I would need to go to actually pursue some legal remedy I'm unlikely to do it regardless of whether it is legal because it is expensive, if not in money but time.
In the case of copyright lawsuits they are notoriously expensive and long so even if a court would eventually deem it fine, why take the chance.
Ignoring the legal or ethical concerns. Let’s say we live in a world where the cost of copying code is so close to zero that it’s indistinguishable from a world without copyright.
Anything you put out can and will be used by whatever giant company wants to use it with no attribution whatsoever.
Doesn’t that massively reduce the incentive to release the source of anything ever?
No, because (most) people don't work on OSS for vanity, they do it to help other people, whether it's individuals or groups of individuals, ie corporations.
It's the same question as, if an AI can generate "art", or photographers can capture a scene better than any (realistic) painter, then will people still create art? Obviously yes, and we see it of course after Stable Diffusion was released three years ago, people are still creating.
Yes, and it reduces the incentives to release binaries too. Such a world will be populated by almost entirely SaaS, which can still compete on freedom.
Oracle had it's day in court with Google over the Java APIs. Reimplementing APIs can be done without copyright infringement, but Oracle must have tried to find real infringement during discovery.
In this case, we could theoretically prove that the new chardet is a clean reimplementation. Blanchard can provide all of the prompts necessary to re-implement again, and for the cost of the tokens anyone can reproduce the results.
If you only stick to the API and ignore the implementation, it is not Mickey Mouse any more but a rodent. If it was just a clone it wouldn't be 50x as fast. Nevertheless, APIs apparently can be copyrightable. I generally disagree with this; it's how PC compatibles took off, giving consumers better options.
The article is poorly written. Blanchard was a chardet maintainer for years. Of course he had looked at it's code!
What he claimed, and what was interesting, was that Claude didn't look at the code, only the API and the test suite. The new implementation is all Claude. And the implementation is different enough to be considered original, completely different structure, design, and hey, a 48x improvement in performance! It's just API-compatible with the original. Which as per the Google Vs oracle 2021 decision is to be considered fair use.
What if we said that generative AI output is simply not copyrightable. Anything an AI spits out would automatically be public domain, except in cases where the output directly infringes the rights of an existing work.
This would make it so relicensing with AI rewrites is essentially impossible unless your goal is to transition the work to be truly public domain.
I think this also helps somewhat with the ethical quandary of these models being trained on public data while contributing nothing of value back to the public, and disincentivize the production of slop for profit.
> No Copyright Protection for AI-Assisted Creations: Thaler v. Perlmutter
> A recent key judicial development on this topic occurred when the U.S. Supreme Court declined to review the case of Thaler v. Perlmutter on March 2, 2026, effectively upholding lower court rulings that AI-generated works lacking human authorship are not eligible for copyright protection under U.S. law
> > A recent key judicial development on this topic occurred when the U.S. Supreme Court declined to review the case of Thaler v. Perlmutter on March 2, 2026, effectively upholding lower court rulings that AI-generated works lacking human authorship are not eligible for copyright protection under U.S. law
This was AI summary? Those words were not in the article.
The courts said Thaler could not have copyright because he refused to list himself as an author.
In the corporate world, we've started using reimplementation as a way to access tooling that security won't authorize.
Sec has a deny by default policy. Eng has a use-more-AI policy. Any code written in-house is accepted by default. You can see where this is going.
We've been using AI to reimplement tooling that security won't approve. The incentives conspired in the worst outcome, yet here we are. If you want a different outcome, you need to create different incentives.
If Blanchard is claiming not to have been substantively involved in the creation of the new implementation of chardet (i.e. "Claude did it"), then the new implementation is machine generated, and in the USA cannot be copyright and thus cannot be licensed.
If he is claiming to have been somehow substantively "enough" involved to make the code copyrightable, then his own familiarity with the previous LGPL implementation makes the new one almost certainly a derivative of the original.
There's a Japanese version of that page, written in classical text writing direction, in columns. Which is cool. Makes me wonder, though - how readable is it with so many English loanwords which should be rotated sideways to fit into columns?
Total digression but yeah, that layout is stupid and the way those words are dropped in using Romaji makes no sense. That's not how Japanese people lay out pages on the web. In fact I don't think I've ever seen a Japanese web page laid out like a book like this, and in general I'd expect the English proper nouns and words that don't have obvious translations to get transliterated into Katakana. Smells like automatic conversion added by someone not really familiar with common usage.
This is only worth arguing about because software has value. Putting this in context of a world where the cost of writing code is trending to 0, there are two obvious futures:
1. The cost continues to trend to 0, and _all_ software loses value and becomes immediately replaceable. In this world, proprietary, copyleft and permissive licenses do not matter, as I can simply have my AI reimplement whatever I want and not distribute it at all.
2. The coding cost reduction is all some temporary mirage, to be ended soon by drying VC money/rising inference costs, regulatory barriers, etc. In that world we should be reimplementing everything we can as copyleft while the inferencing is good.
There’s an other option. The cost of copying existing software trends to 0, but the cost of writing new software stays far enough above 0 that it is still relatively expensive.
There was a recent ruling that LLM output is inherently public domain (presumably unless it infringes some existing copyright). In which case it's not possible to use them to "reimplement everything we can as copyleft".
it's more complicated, the ruling was that AI can't be an author and the thing in question is (de-facto) public domain because it has no author in context of the "dev" claim it was fully build by AI
but AI assisted code has an author and claiming it's AI assisted even if it is fully AI build is trivial (if you don't make it public that you didn't do anything)
also some countries have laws which treat it like a tool in the sense that the one who used it is the author by default AFIK
The article is proceeding from the premise that a reimplementation is legal (but evil). To help my understanding of your comment, do you mean:
1. An LLM recreating a piece of software violates its copyright and is illegal, in which case LLM output can never be legally used because someone somewhere probably has a copyright on some portion of any software that an LLM could write.
2. You read my example as "copying a project without distributing it", vs. "having an LLM write the same functionality just for me"
This article is setting up a bit of a moving target. Legal vs legitimate is at least only a single vague question to be defined but then the target changes to “socially legitimate” defined only indirectly by way of example, like aggressive tax avoidance as “antisocial”— and while I tend to agree with that characterization my agreement is predicated on a layering of other principals.
The fundamental problem is that once you take something outside the realm of law and rule of law in its many facets as the legitimizing principal, you have to go a whole lot further to be coherent and consistent.
You can’t just leave things floating in a few ambiguous things you don’t like and feel “off” to you in some way- not if you’re trying to bring some clarity to your own thoughts, much less others. You don’t have to land on a conclusion either. By all means chew over things, but once you try to settle, things fall apart if you haven’t done the harder work of replacing the framework of law with that of another conceptual structure.
You need to at least be asking “to what ends? What purpose is served by the rule?” Otherwise you’re stuck in things where half the time you end up arguing backwards in ways that put purpose serving rules, the maintenance of the rule with justifications ever further afield pulled in when the rule is questioned and edge cases reached. If you’re asking, essentially, “is the spirit of the rule still there?” You’ve got to stop and fill in what that spirit is or you or people that want to control you or have an agenda will sweep in with their own language and fill the void to their own ends.
Surprised they don't mention Google LLC v. Oracle America, Inc. Seems a bit myopic to condone the general legality while arguing "you can only use it how I like it".
It also doesn't talk about the far more interesting philosophical queston. Does what Blanchard did cover ALL implementations from Claude? What if anyone did exactly what he did, feed it the test cases and say "re-implement from scratch", ostensibly one would expect the results to be largely similar (technically under the right conditions deterministically similar)
could you then fork the project under your own name and a commercial license? when you use an LLM like this, to basically do what anyone else could ask it to do how do you attach any license to it? Is it first come first serve?
If an agent is acting mostly on its own it feels like if you found a copy of Harry Potter in the fictional library of Babel, you didn't write it, just found it amongst the infinite library, but if you found it first could you block everyone else that stumbles on a near-identical copy elsewhere in the library? or does each found copy represent a "Re-implementation" that could be individually copyrighted?
It should be noted that the Rust community is also guilty of something similar. That is, porting old GPL programs, typically written in C, to Rust and relicensing them as MIT.
Broadly speaking, the “freedom of users” is often protected by competition from competing alternatives. The GNU command line tools were replacements for system utilities. Linux was was a replacement for other Unix kernels. People chose to install them instead of proprietary alternatives. Was it due to ideology or lower cost or more features? All of the above. Different users have different motivations.
Copyleft could be seen as an attempt to give Free Software an edge in this competition for users, to counter the increased resources that proprietary systems can often draw on. I think success has been mixed. Sure, Linux won on the server. Open source won for libraries downloaded by language-specific package managers. But there’s a long tail of GPL apps that are not really all that appealing, compared to all the proprietary apps available from app stores.
But if reimplementing software is easy, there’s just going to be a lot more competition from both proprietary and open source software. Software that you can download for free that has better features and is more user-friendly is going to have an advantage.
With coding agents, it’s likely that you’ll be able to modify apps to your own needs more easily, too. Perhaps plugin systems and an AI that can write plugins for you will become the norm?
Why are people even having problems with sharing their changes to begin with? Just publishing it somewhere does not seem too expensive. The risk of accidentally including stuff that is not supposed to become public? Or are people regularly completely changing codebases and do not want to make the effort freely available, maybe especially to competitors? I would have assumed that the common case is adding a missing feature here, tweaking something there, if you turn the entire thing on its head, why not have your own alternative solution from scratch?
You can't put a copyright and MIT license on something you generated with AI. It is derived from the work of many unknown, uncredited authors.
Think about it; the license says that copies of the work must be reproduced with the copyright notice and licensing clauses intact. Why would anyone obey that, knowing it came from AI?
Countless instances of such licenses were ignored in the training data.
IMHO, the API and Test Suite, particularly the latter, define the contract of the functional definition of the software. It almost doesn't matter what that definition looks like so long as it conforms to the contract.
There was an issue where Google did something similar with the JVM, and ultimately it came down to whether or not Oracle owned the copyright to the header files containing the API. It went all the way to the US supreme court, and they ruled in Google's favour; finding that the API wasn't the implementation, and that the amount of shared code was so minimal as to be irrelevant.
They didn't anticipate that in less than half a decade we'd have technology that could _rapidly_ reimplement software given a strong functional definition and contract enforcing test suite.
"Antirez closes his careful legal analysis as though it settles the matter. Ronacher acknowledges that “there is an obvious moral question here, but that isn't necessarily what I'm interested in.” Both pieces treat legal permissibility as a proxy for social legitimacy. "
This whole article is just complaining that other people didn't have the discussion he wanted.
Ronacher even acknowledged that it's a different discussion, and not one they were trying to have at the moment.
If you want to have it, have it. Don't blast others for not having it for you.
Having this discussion involves blasting others for not considering it. Consider the rest of the paragraph you quoted:
> But law only says what conduct it will not prevent—it does not certify that conduct as right. Aggressive tax minimization that never crosses into illegality may still be widely regarded as antisocial. A pharmaceutical company that legally acquires a patent on a long-generic drug and raises the price a hundredfold has not done something legal and therefore fine. Legality is a necessary condition; it is not a sufficient one.
> If source code can now be generated from a specification, the specification is where the essential intellectual content of a GPL project resides. Blanchard's own claim—that he worked only from the test suite and API without reading the source—is, paradoxically, an argument for protecting that test suite and API specification under copyleft terms.
This is an interesting reversal in itself. If you make the specification protected under copyright, then the whole practice of clean room implementations is invalid.
Not a lawyer, but my understanding is: In theory, copyright only protects the creative expression of source code; this is the point of the "clean room" dance, that you're keeping only the functional behavior (not protected by copyright). Patents are, of course, an entirely different can of worms. So using an LLM to strip all of the "creative expression" out of source code but create the same functionality feels like it could be equivalent enough.
I like the article's point of legal vs. legitimate here, though; copyright is actually something of a strange animal to use to protect source code, it was just the most convenient pre-existing framework to shove it in.
> When GNU reimplemented the UNIX userspace, the vector ran from proprietary to free. Stallman was using the limits of copyright law to turn proprietary software into free software. […] The vector in the chardet case runs the other way.
That’s just your subjective opinion which many other people would disagree. I bet Armin Ronacher would agree that an MIT licensed library is even freer than an LGPL licensed library. To them, the vector is running from free to freer.
If you decide to improve it in any way to fit your needs you can merely tell your own AI to re-implement it with your changes. Then it's proprietary to you.
I feel like the licenses that suffer the most isn't the GPL, but the ones like SSPL. If your code can be re-implemented easily and legally by AWS using an LLM, why risk publishing it?
It does feel like open source is about to change. My hunch is that commercial open source (beyond the consultation model) risks disappearing. Though I'd be happy to be proven wrong.
It's clear that we're entering a new era of copyright _expectations_ (whether we get new _legislation_ is different), but for now realise this: the people like me who like copyleft can do this too. We can take software we like, point an agent at it, and tell it to make a new version with the AGPL3.0-or-later badge on the front.
I don't think this part is correct: "If you distribute modified code, or offer it as a networked service, you must make the source available under the same terms."
One of the things that irks me about this whole thing is, if it’s so clean room and distinct, why make the changes to the existing project? Why not make an entirely new library?
The answer to that, I think, is that the authors wanted to squat an existing successful project and gain a platform from it. Hence we have news cycle discussing it.
Nobody cares about a new library using AI, but squash an existing one with this stuff, and you get attention. It’s the reputation, the GitHub stars, whatever
I mean, Blanchard was the longtime maintainer of chardet already, and had wanted to relicense it for years. So I think that complicates your picture of "squatting an existing successful project".
Honestly it's a weird test case for this sort of thing. I don't think you'd see an equivalent in most open source projects.
Imagine if the author has his way, and when we have AI write software, it becomes legally under the license of some other sufficiently similar piece of software. Which may or may not be proprietary. "I see you have generated a todo app very similar to Todoist. So they now own it." That does not seem like a good path either for open source software or for opening up the benefits of AI generated software.
What if someone doesn't declare that it has been reimplemented using an LLM? Isn't it enough to simply declare that you have reimplemented the software without using an LLM? Good luck proving that in court...
One thing is certain, however: copyleft licenses will disappear: If I can't control the redistribution of my code (through a GPL or similar license), I choose to develop it in closed source.
If the model wasn't trained on copyleft, if he didn't use a copyleft test suite and if he wasn't the maintainer for years. Clearly the intent here is copyright infringement.
If you have software your testsuite should be your testsuite, you do dev with a testsuite and then mit without releasing one. Depending on the test-suite it may break clean room rules, especially for ttd codebases.
I think what is happening is the collapse of the “greater good”. Open source is dependent upon providing information for the greater good and general benefit of its readers. However now that no one is reading anything, its purpose is for the great good of the most clever or most convincing or richest harvester.
I think AI is very much eroding the legitimacy of copyright - at least to software, which is long been questioned since it's more like math than creative expression.
I think the industry will realize that it made a huge mistake by leaning on copyright for protection rather than on patents.
Probably a wiser approach is to consider different times require different measures (in general!).
I did not study in detail if copyright "has always been nonsense", but I do agree that nowadays some of the copyright regulations are nonsense (for example the very long duration of life + 70 years)
> Ronacher notes this as an irony and moves on. But the irony cuts deeper than he lets on. Next.js is MIT licensed. Cloudflare's vinext did not violate any license—it did exactly what Ronacher calls a contribution to the culture of openness, applied to a permissively licensed codebase. Vercel's reaction had nothing to do with license infringement; it was purely competitive and territorial. The implicit position is: reimplementing GPL software as MIT is a victory for sharing, but having our own MIT software reimplemented by a competitor is cause for outrage. This is what the claim that permissive licensing is “more share-friendly” than copyleft looks like in practice. The spirit of sharing, it turns out, runs in one direction only: outward from oneself.
This argument makes no sense. Are they arguing that because Vercel, specifically, had this attitude, this is an attitude necessitated by AI, reimplementation, and those who are in favor of it towards more permissive licenses? That certainly doesn't seem to be an accurate way to summarize what antirez or Ronacher believe. In fact, under the legal and ethical frameworks (respectively) that those two put forward, Vercel has no right to claim that position and no way to enforce it, so it seems very strange to me to even assert that this sort of thing would be the practical result of AI reimplementations. This seems to just be pointing towards the hypocrisy of one particular company, and assuming that this would be the inevitable universal, attitude, and result when there's no evidence to think so.
It's ironic, because antirez actually literally addresses this specific argument. They completely miss the fact that a lot of his blog post is not actually just about legal but also about ethical matters. Specifically, the idea he puts forward is that yes, corporations can do these kinds of rewrites now, but they always had the resources and manpower to do so anyway. What's different now is that individuals can do this kind of rewrites when they never have the ability to do so before, and the vector of such a rewrite can be from a permissive to copyleft or even from decompile the proprietary to permissive or copyleft. The fact that it hasn't been so far is a more a factor of the fact that most people really hate copyleft and find an annoying and it's been losing traction and developer mind share for decades, not that this tactic can't be used that way. I think that's actually one of the big points he's trying to make with his GNU comparison — not just that if it was legal for GNU to do it, then it's legal for you to do with AI, and not even just the fundamental libertarian ethical axiom (that I agree with for the most part) that it should remain legal to do such a rewrite in either direction because in terms of the fundamental axioms that we enforce with violence in our society, there should be a level playing field where we look at the action itself and not just whether we like or dislike the consequences, but specifically the fact that if GNU did it once with the ability to rewrite things, it can be done again, even in the same direction, it now even more easily using AI.
> They completely miss the fact that a lot of his blog post is not actually just about legal but also about ethical matters.
Honestly I was confused about the summarization of my blog post into just a legal matter as well. I hope my blog post will be able to flash at least a short time in the HN front page so that the actual arguments it contain will get a bit more exposure.
I'm failing to see what in the quoted text you took to be about AI rewrites specifically? It just reads as a slightly catty aside about the social reaction of rewrites in general (by implying the one example is generalizable.)
I think we're going one step too far even, AI itself is a gray area and how can they guarantee it was trained legally or if it's even legal what they're doing and how can they assert that the input training data didn't contain any copyrighted data.
Google already spent billions of dollars and decades of lawyer hours proving it out as fair use. The legal challenges we see now are the dying convulsions of an already broken system of publishers and IP hoarders using every resource at their disposal to manipulate authors and creators and the public into thinking that there's any legitimacy or value underlying modern copyright law.
AI will destroy the current paradigm, completely and utterly, and there's nothing they can do to stop it. It's unclear if they can even slow it, and that's a good thing.
We will be forced to legislate a modern, digital oriented copyright system that's fair and compatible with AI. If producing any software becomes a matter of asking a machine to produce it - if things like AI native operating systems come about, where apps and media are generated on demand, with protocols as backbone, and each device is just generating its own scaffolding around the protocols - then nearly none of modern licensing, copyright, software patents, or IP conventions make any sense whatsoever.
You can't have horse and buggy traffic conventions for airplanes. We're moving in to a whole new paradigm, and maybe we can get legislation that actually benefits society and individuals, instead of propping up massive corporations and making lawyers rich.
Google has cut out some very specific ruling that have nothing to do with modern AI. These systems are just a really slow/lossy git clone, current law has no trouble with it, it's broadly illegal.
If corporations are allowed to launder someone else work as their own people will simply stop working and just start endlessly remixing a la popular music.
I believe it is a narrow view of the situation. If we take a look into the history, into the reasons for inventing GPL, we'll see that it was an attempt to fight copyrights with copyrights. The very name 'copyleft' is trying to convey the idea.
What AI are eroding is copyright. You can re-implement not just a GPL program, but to reverse engineer and re-implement a closed source program too, people have demonstrated it already, there were stories here on HN about it.
AI is eroding copyright, so there may no longer be a need for the GPL. GNU should stop and rethink its stance, chuck away the GPL as the main tool to fight evil software corporations and embrace LLM as the main weapon.
> LLM as the main weapon
LLM's - to date - seem to require massive capital expenditures to have the highest quality ones, which is a monumental shift in power towards mega corporations and away from the world of open source where you could do innovative work on your own computer running Linux or FreeBSD or some other open OS.
I don't think that's an exciting idea for the Free Software Foundation.
Perhaps with time we'll be able to run local ones that are 'good enough', but we're not there yet.
There's also an ethical/moral question that these things have been trained on millions of hours of people's volunteer work and the benefits of that are going to accrue to the mega corporations.
Edit: I guess the conclusion I come to is that LLM's are good for 'getting things done', but the context in which they are operating is one where the balance of power is heavily tilted towards capital, and open source is perhaps less interesting to participate in if the machines are just going to slurp it up and people don't have to respect the license or even acknowledge your work.
> LLM's - to date - seem to require massive capital expenditures to have the highest quality ones, which is a monumental shift in power towards mega corporations and away from the world of open source
Yeah, a bit of a conundrum. But I don't think that fighting for copyright now can bring any benefits for FOSS. GNU should bring Stallman back and see whether he can come with any new ideas and a new strategy. Alternatively they could try without Stallman. But the point is: they should stop and think again. Maybe they will find a way forward, maybe they won't but it means that either they could continue their fight for a freedom meaningfully, or they could just stop fighting and find some other things to do. Both options are better then fighting for copyright.
> There's also an ethical/moral question that these things have been trained on millions of hours of people's volunteer work and the benefits of that are going to accrue to the mega corporations.
I want a clarify this statement a bit. The thing with LLM relying on work of others are not against GPU philosophy as I understand it: algorithms have to be free. Nothing wrong with training LLMs on them or on programs implementing them. Nothing wrong with using these LLMs to write new (free) programs. What is wrong are corporations reaping all the benefits now and locking down new algorithms later.
I think it is important, because copyright is deemed to be an ethical thing by many (I think for most people it is just a deduction: abiding the law is ethical, therefore copyright is ethical), but not for GNU.
Is massive capital expenditure not also required to enforce the GPL? If some company steals your GPLed code and doesn't follow the license, you will have to sue them and somebody will have to pay the lawyers.
> Is massive capital expenditure not also required to enforce the GPL?
It's nowhere near the order of magnitude of the kind of spending they're sinking into LLM's. The FSF and other groups were reasonably successful at enforcing the GPL, operating on a budget 1000's of times smaller than that of AI companies.
> LLM's - to date - seem to require massive capital expenditures to have the highest quality ones
There are near-SOTA LLM's available under permissive licenses. Even running them doesn't require prohibitive expenses on hardware unless you insist on realtime use.
> There's also an ethical/moral question that these things have been trained on millions of hours of people's volunteer work and the benefits of that are going to accrue to the mega corporations.
This was already the case and it just got worse, not better.
At a certain point, I think we had reached a kind of equilibrium where some corporations were decent open source citizens. They understood that they could open source things like infrastructure or libraries and keep their 'crown jewels' closed. And while Stallman types might not have been happy with that, it seemed to work out for people.
Now they've just hoovered up all the free stuff into machines that can mix it up enough to spit it out in a way that doesn't even require attribution, and you have to pay to use their machine.
AI essentially gatekeeps all of open source to companies to pluck from to their hearts content. And individual contributors using these tools and freely mixing it with their own - usual minor - contributions are another step of whitewashing because they're definitely not going to own up to writing only 5% of the stuff they got paid for.
Before we had RedHat and Ubuntu, who at least were contributing back, now we have Microsoft, Anthropic and OpenAI who are racing to lock the barn door around their new captive sheep. It's just a massive IP laundromat.
Copyleft is a mirror of copyright, not a way to fight copyright. It grants rights to the consumer where copyright grants rights to the creator. Importantly, it gives the end-user the right to modify the software running on their devices.
Unfortunately, there are cases where you simply can't just "re-implement" something. E.g., because doing so requires access to restricted tools, keys, or proprietary specifications.
These are words of Stallman:
"So, I looked for a way to stop that from happening. The method I came up with is called “copyleft.” It's called copyleft because it's sort of like taking copyright and flipping it over. [Laughter] Legally, copyleft works based on copyright. We use the existing copyright law, but we use it to achieve a very different goal."
https://writings.hongminhee.org/2026/03/legal-vs-legitimate/
> flipping it over.
i.e. mirroring it
> use it to achieve a very different goal."
"very different goal" isn't the same as "fundamentally destroying copyright"
the very different goal include to protect public code to stay public, be properly attributed, prevent companies from just "sizing" , motivate other to make their code public too etc.
and even if his goals where not like that, it wouldn't make a difference as this is what many people try to archive with using such licenses
this kind of AI usage is very much not in line with this goals,
and in general way cheaper to do software cloning isn't sufficient to fix many of the issues the FOSS movement tried to fix, especially not when looking at the current ecosystem most people are interacting with (i.e. Phones)
---
("sizing"): As in the typical MS embrace, extend and extinguish strategy of first embracing the code then giving it proprietary but available extensions/changes/bug fixes/security patches to then make them no longer available if you don't pay them/play by their rules.
---
Through in the end using AI as a "fancy complicated" photocopier for code is as much removing copyright as using a photocopier for code would. It doesn't matter if you use the photocopier blind folded and never looked at the thing you copied.
That’s not a rebuttal of the OP’s point. None of that says anything about fighting copyright. It literally says he flipped it which is wha the OP said when they said it’s a mirror.
> It grants rights to the consumer where copyright grants rights to the creator.
It also grants one major right/feature to the creator, the ability to spread their work while keeping it as open as they intend.
> AI is eroding copyright, so there may no longer be a need for the GPL. GNU should stop and rethink its stance, chuck away the GPL as the main tool to fight evil software corporations and embrace LLM as the main weapon.
Is this LLM thing freely available or is it owned and controlled by these companies? Are we going to rent the tools to fight "evil software corporations"?
easy, we ask Claude to write an open-source freely-available version of Claude with equal or better capabilities.
Its purpose "if you run the software you should be able to inspect and modify that software, and to share those modifications with your peers" not explicitly resist copyright. Yes copyright is bad in that it often prevents one from doing that, but it is not the purpose of the GPL to dismantle copyright.
Reducing it to "well you can clone the proprietary software you're forced to use by LLM" is really missing the soul of the GPL.
If not for copyright, you could always do that and copyleft wouldn't be needed.
> we'll see that it was an attempt to fight copyrights with copyrights
it's not that simple
yes, GPLs origins have the idea of "everyone should be able to use"
but it also is about attribution the original author
and making sure people can't just de-facto "size public goods"
the kind of AI usage is removing attribution and is often sizing public goods in a way far worse then most companies which just ignored the license did
so today there is more need then ever in the last few decades for GPL like licenses
You've said "size" twice in comments, did you mean "seize"?
So not only are we moving goalposts here, but we've decided the GNU team should join the other team? I don't understand how GNU would see mass model LLM training as anything but the most flagrant violations of their ethos. LLM labs, in their view, would be among the most evil software corporations to have ever existed.
That's naive. Copyright doesn't just apply to software. There already have been countless lawsuits about copying music long before the term "open source" was invented. No, changing the lyrics a bit doesn't circumvent copyright. Nor does translating a Stephen King novel to German and switching the names of the places and characters.
A court ordered the first Nosferatu movie to be destroyed because it had too many similarities to Dracula. Despite the fact that the movie makes rather large deviations from the original.
If Claude was indeed asked to reimplement the existing codebase, just in Rust and a bit optimized, that could well be a copyright violation. Just like rephrasing A Song ot Ice and Fire a bit, and switching to a different language, doesn't remove its copyright.
Claude was asked to implement a public API, not an entire codebase. The definition of a public API is largely functional; even in an unusually complex case like the Java standard facilities (which are unusually creative even in the structure and organization of the API itself) the reimplementation by Google was found to be fair use.
> Claude was asked to implement a public API, not an entire codebase.
Allegedly. There have been several people who doubted this story. So how to find out who is right? Well, just let Claude compare the sources. Coincidentally, Claude Opus 4.6 doesn't just score 75.6% on SWE-bench Verified but also 90.2% on BigLaw Bench.
It's like our copyright lawyer is conveniently also a developer. And possibly identical to the AI that carried out the rewrite/reimplemention in question in the first place.
> What AI are eroding is copyright.
At the moment it's people that are eroding copyright. E.g. in this case someone did something.
"AI" didn't have a brain, woke up and suddenly decided to do it.
Realistically nothing to do with AI. Having a gun doesn't mean you randomly shoot.
While I personally agree with you, Richard Stallman (the creator of the GPL) does not. He has always advocated in favor of strong copyright protection, because the foundation of the GPL is the monopoly power granted by copyright. The problem that the GPL is intended to solve is proprietary software.
Generative models (AI) are not really eroding copyright. They are calling its bluff. The very notion of intellectual property depends on a property line: some arbitrary boundary where the property begins and ends. Generative models blur that line, making it impractical to distinguish which property belongs to whom.
Ironically, these models are made by giant monopolistic corporations whose wealth is quite literally a market valuation (stock price) of their copyrights! If generative models ever become good enough to reimplement CUDA, what value will NVIDIA have left?
The reality is that generative models are nowhere near good enough to actually call the bluff. Copyright is still the winning hand, and that is likely to continue, particularly while IP holders are the primary authors of law.
---
This whole situation is missing the forest for the trees. Intellectual Property is bullshit. A system predicated on monopoly power can only result in consolidated wealth driving the consolidation of power; which is precisely what has happened. The words "starving artist" ring every bit as familiar today as any time in history. Copyright has utterly failed the very goals it was explicitly written with.
It isn't the GPL that needs changing. So long as a system of copyright rules the land, copyleft is the best way to participate. What we really need is a cohesive political movement against monopoly power; one that isn't conveniently ignorant of copyright as its most significant source.
> Blanchard's account is that he never looked at the existing source code directly. He fed only the API and the test suite to Claude and asked it to reimplement the library from scratch
This feels sort of like saying "I just blindly threw paint at that canvas on the wall and it came out in the shape of Mickey Mouse, and so it can't be copyright infringement because it was created without the use of my knowledge of Micky Mouse"
Blanchard is, of course, familiar with the source code, he's been its maintainer for years. The premise is that he prompted Claude to reimplement it, without using his own knowledge of it to direct or steer.
> Blanchard is, of course, familiar with the source code, he's been its maintainer for years.
I would argue it's irrelevant if they looked or didn't look at the code. As well as weather he was or wasn't familiar with it.
What matters is, that they feed to original code into a tool which they setup to make a copy of it. How that tool works doesn't really matter. Neither does it make a difference if you obfuscate that it's an copy.
If I blindfold myself when making copies of books with a book scanner + printer I'm still engaging in copyright infringement.
If AI is a tool, that should hold.
If it isn't "just" a tool, then it did engage in copyright infringement (as it created the new output side by side with the original) in the same way an employee might do so on command of their boss. Which still makes the boss/company liable for copyright infringement and in general just because you weren't the one who created an infringing product doesn't mean you aren't more or less as liable of distributing it, as if you had done so.
> If it isn't "just" a tool, then it did engage in copyright infringement
Copyright infringement is a thing humans do. It's not a human.
Just like how the photos taken by a monkey with a camera have no copyright. Human law binds humans.
Correct. The human who shares the copy is the one who engages in copyright infringement.
So, let's say that rather than actually touching any copyrighted material, a human merely tells an AI about how to go onto the internet and find copyrighted material, download it, and ingest it for training. The AI, fully autonomously, does so, and after training itself on the material deletes it so no human ever downloads, consumes, or shares it.
If we are saying AI is "more than a tool", which seems to be the case courts are leaning since they've ruled AI output without direct human involvement is not copyrightable[0], then the above seems like it would be entirely legal.
[0] https://www.copyright.gov/newsnet/2025/1060.html
if the actual text of the code isn't the same or obviously derivative, copyright doesn't apply at all.
Copyright protects even very abstract aspects of human creative expression, not just the specific form in which it is originally expressed. If you translate a book into another language, or turn it into a silent movie, none of the actual text may survive, but the story itself remains covered by the original copyright.
So when you clone the behavior of a program like chardet without referencing the original source code except by executing it to make sure your clone produces exactly the same output, you may still be infringing its copyright if that output reflects creative choices made in the design of chardet that aren't fully determined by the functional purpose of the program.
If you pirate a movie and reencode it, does that apply as well? You can still watch the movie and it is “obviously” the same movie. Here you can use the program and it is, to the user, also the same.
What does derivative mean here? Because IMO it means that the existing work was used as input. So if you used a LLM and it was trained on the existing work, that's a derivative work. If you rot13 encode something as input, so you can't personally read it, and then a device decides to rot13 on it again and output it, that's a derivative work.
In order for it to be creatively derivative you would need to copy the structure, logic, organization, and sequence of operations not just reimplement the functionality. It is pretty clear in this case that wasn't done.
As a cynical person I assume all the frontier LLMs were trained on datasets that include every open source project, but as a thought experiment, if an LLM was trained on a dataset that included every open source project _execept_ chardet, do you think said LLM would still be able to easily implement something very similar?
There is no doubt in my mind that it could still do it.
Of course, the problem with this interpretation is that all modern LLMs are derivatives from huge amounts of text under completely different licenses, including "All rights reserved", and therefore can not be used for any purpose.
I'm not sure how you square the circle of "it's alright to use the LLM to write code, unless the code is a rewrite of an open source project to change its license".
> Because IMO it means that the existing work was used as input
That's your opinion (since you said "IMO"), not the actual legal definition.
LLMs do not encode nor encrypt their training data. The fact they can recite training data is a defect not a default. You can understand this more simply by calculating the model size as an inverse of a fantasy compression algorithm that is 50% better than SOTA. You'll find you'd still be missing 80-90% of the training data even if it were as much of a stochastic parrot as you may be implying. The outputs of AI are not derivative just because they saw training data including the original library.
Then onto prompting: 'He fed only the API and (his) test suite to Claude'
This is Google v Oracle all over again - are APIs copyrightable?
> This is Google v Oracle all over again - are APIs copyrightable?
Yes this is the best way to ask the question. If I take a public facing API and reimplement everything, whether it's by human or machine, it should be sufficient. After all, that's what Google did, and it's not like their engineers never read a single line of the Java source code. Even in "clean room" implementations, a human might still have remembered or recalled a previous implementation of some function they had encountered before.
See also: https://monolith.sourceforge.net/, which seeks to ask the question:
> But how far away from direct and explicit representations do we have to go before copyright no longer applies?
I just don't see how it's relevant whether he did look or didn't. In my opinion, it's not just legally valid to make a re-implementation of something if you've seen the code as long as it doesn't copy expressive elements. I think it's also ethically fine as well to use source code as a reference for re-implementing something as long as it doesn't turn into an exact translation.
Right. The alternative is that we reward Dan for his 14 years of volunteer maintenance of a project... by banning him from working on anything similar under a different license for the rest of his life.
It's actually not legally fine, or at least it's extremely dangerous. Projects that re-implement APIs presented by extremely litigious companies specifically do not allow people who, for instance, have seen the proprietary source code to then work on the project.
I don't think fear or legal action makes it illegal.
If I know it is legal to make a turn at a red light. And I know a court will uphold that I was in the right but a police officer will fine me regardless and I would need to go to actually pursue some legal remedy I'm unlikely to do it regardless of whether it is legal because it is expensive, if not in money but time.
In the case of copyright lawsuits they are notoriously expensive and long so even if a court would eventually deem it fine, why take the chance.
My understanding is that that is a maximalist position for the avoidance of risk, and is sufficient but probably not necessary.
Ignoring the legal or ethical concerns. Let’s say we live in a world where the cost of copying code is so close to zero that it’s indistinguishable from a world without copyright.
Anything you put out can and will be used by whatever giant company wants to use it with no attribution whatsoever.
Doesn’t that massively reduce the incentive to release the source of anything ever?
No, because (most) people don't work on OSS for vanity, they do it to help other people, whether it's individuals or groups of individuals, ie corporations.
It's the same question as, if an AI can generate "art", or photographers can capture a scene better than any (realistic) painter, then will people still create art? Obviously yes, and we see it of course after Stable Diffusion was released three years ago, people are still creating.
Yes, and it reduces the incentives to release binaries too. Such a world will be populated by almost entirely SaaS, which can still compete on freedom.
Most commercial software that I've used has the model of a legal moat around a pretty crappy database schema.
The non IP protection has largely been in the effort involved in replicating an application's behavior and that effort is dropping precipitously.
Oracle had it's day in court with Google over the Java APIs. Reimplementing APIs can be done without copyright infringement, but Oracle must have tried to find real infringement during discovery.
In this case, we could theoretically prove that the new chardet is a clean reimplementation. Blanchard can provide all of the prompts necessary to re-implement again, and for the cost of the tokens anyone can reproduce the results.
Can anyone find the actual quote where Blanchard said this?
My understanding was that his claim was that Claude was not looking at the existing source code while writing it.
Conveniently ignoring the likelihood that Claude had been trained on the freely accessible source code.
Does he have access to Claude's training data? How can he claim Claude wasn't trained on the original code?
Isn't this a red herring? An API definition is fair use under Google v. Oracle, but the test suite is definitely copyrightable code!
If you only stick to the API and ignore the implementation, it is not Mickey Mouse any more but a rodent. If it was just a clone it wouldn't be 50x as fast. Nevertheless, APIs apparently can be copyrightable. I generally disagree with this; it's how PC compatibles took off, giving consumers better options.
Wait what, didn't oracle lose the case against Google? Have I been living in an alternate reality where API compatibility is fair use?
> This feels sort of like saying "I just blindly threw paint at that canvas on the wall and
> He fed only the API and the test suite to Claude and asked it
Difference being Claude looked; so not blind. The equivalent is more like I blindly took a photo of it and then used that to...
Technically did look.
The article is poorly written. Blanchard was a chardet maintainer for years. Of course he had looked at it's code!
What he claimed, and what was interesting, was that Claude didn't look at the code, only the API and the test suite. The new implementation is all Claude. And the implementation is different enough to be considered original, completely different structure, design, and hey, a 48x improvement in performance! It's just API-compatible with the original. Which as per the Google Vs oracle 2021 decision is to be considered fair use.
did he claim that Claude wasn't trained on the original? Or just that he didn't personally provide Claude with a copy?
I recon the latter, how would he know what was in Claude's training data?
> What he claimed, and what was interesting, was that Claude didn't look at the code
Who opened the PR? Who co-authored the commits? It's clearly on Github.
> Blanchard was a chardet maintainer for years. Of course he had looked at its code!
So there you have it. If he looked, he co-authored then there's that.
If I put my signature on Picasso painting, it doesn't make me co-author of said painting.
Blanchard is very clear that he didn't write a single line of code. He isn't an author, he isn't a co-author.
Signing GitHub commit doesn't change that.
> Blanchard is very clear that he didn't write a single line of code
He used Claude to write it. Difference? The fact that I write on the notepad vs printed it out = I didn't do it?
> Signing GitHub commit doesn't change that.
That's the equivalent of me saying I didn't kill anyone. The fingerprints on the knife doesn't change that.
I'll take a commit authored by someone else and then git amend the author to myself, did I write that commit then? By your logic I did apparently.
What if we said that generative AI output is simply not copyrightable. Anything an AI spits out would automatically be public domain, except in cases where the output directly infringes the rights of an existing work.
This would make it so relicensing with AI rewrites is essentially impossible unless your goal is to transition the work to be truly public domain.
I think this also helps somewhat with the ethical quandary of these models being trained on public data while contributing nothing of value back to the public, and disincentivize the production of slop for profit.
We did in fact say so.
https://www.carltonfields.com/insights/publications/2025/no-...
> No Copyright Protection for AI-Assisted Creations: Thaler v. Perlmutter
> A recent key judicial development on this topic occurred when the U.S. Supreme Court declined to review the case of Thaler v. Perlmutter on March 2, 2026, effectively upholding lower court rulings that AI-generated works lacking human authorship are not eligible for copyright protection under U.S. law
> > A recent key judicial development on this topic occurred when the U.S. Supreme Court declined to review the case of Thaler v. Perlmutter on March 2, 2026, effectively upholding lower court rulings that AI-generated works lacking human authorship are not eligible for copyright protection under U.S. law
This was AI summary? Those words were not in the article.
The courts said Thaler could not have copyright because he refused to list himself as an author.
> This would make it so relicensing with AI rewrites is essentially impossible unless your goal is to transition the work to be truly public domain.
That's not true at all. Anyone could follow these steps:
1. Have the LLM rewrite GPL code.
2. Do not publish that public domain code. You have no obligation to.
3. Make a few tweaks to that code.
4. Publish a compiled binary/use your code to host a service under a proprietary license of your choice.
In the corporate world, we've started using reimplementation as a way to access tooling that security won't authorize.
Sec has a deny by default policy. Eng has a use-more-AI policy. Any code written in-house is accepted by default. You can see where this is going.
We've been using AI to reimplement tooling that security won't approve. The incentives conspired in the worst outcome, yet here we are. If you want a different outcome, you need to create different incentives.
If Blanchard is claiming not to have been substantively involved in the creation of the new implementation of chardet (i.e. "Claude did it"), then the new implementation is machine generated, and in the USA cannot be copyright and thus cannot be licensed.
If he is claiming to have been somehow substantively "enough" involved to make the code copyrightable, then his own familiarity with the previous LGPL implementation makes the new one almost certainly a derivative of the original.
There's a Japanese version of that page, written in classical text writing direction, in columns. Which is cool. Makes me wonder, though - how readable is it with so many English loanwords which should be rotated sideways to fit into columns?
Total digression but yeah, that layout is stupid and the way those words are dropped in using Romaji makes no sense. That's not how Japanese people lay out pages on the web. In fact I don't think I've ever seen a Japanese web page laid out like a book like this, and in general I'd expect the English proper nouns and words that don't have obvious translations to get transliterated into Katakana. Smells like automatic conversion added by someone not really familiar with common usage.
This is only worth arguing about because software has value. Putting this in context of a world where the cost of writing code is trending to 0, there are two obvious futures:
1. The cost continues to trend to 0, and _all_ software loses value and becomes immediately replaceable. In this world, proprietary, copyleft and permissive licenses do not matter, as I can simply have my AI reimplement whatever I want and not distribute it at all.
2. The coding cost reduction is all some temporary mirage, to be ended soon by drying VC money/rising inference costs, regulatory barriers, etc. In that world we should be reimplementing everything we can as copyleft while the inferencing is good.
There’s an other option. The cost of copying existing software trends to 0, but the cost of writing new software stays far enough above 0 that it is still relatively expensive.
There was a recent ruling that LLM output is inherently public domain (presumably unless it infringes some existing copyright). In which case it's not possible to use them to "reimplement everything we can as copyleft".
it's more complicated, the ruling was that AI can't be an author and the thing in question is (de-facto) public domain because it has no author in context of the "dev" claim it was fully build by AI
but AI assisted code has an author and claiming it's AI assisted even if it is fully AI build is trivial (if you don't make it public that you didn't do anything)
also some countries have laws which treat it like a tool in the sense that the one who used it is the author by default AFIK
The value of software has never been tied to the cost of writing it, even if you don't distribute it your still breaking the law.
The article is proceeding from the premise that a reimplementation is legal (but evil). To help my understanding of your comment, do you mean:
1. An LLM recreating a piece of software violates its copyright and is illegal, in which case LLM output can never be legally used because someone somewhere probably has a copyright on some portion of any software that an LLM could write.
2. You read my example as "copying a project without distributing it", vs. "having an LLM write the same functionality just for me"
This article is setting up a bit of a moving target. Legal vs legitimate is at least only a single vague question to be defined but then the target changes to “socially legitimate” defined only indirectly by way of example, like aggressive tax avoidance as “antisocial”— and while I tend to agree with that characterization my agreement is predicated on a layering of other principals.
The fundamental problem is that once you take something outside the realm of law and rule of law in its many facets as the legitimizing principal, you have to go a whole lot further to be coherent and consistent.
You can’t just leave things floating in a few ambiguous things you don’t like and feel “off” to you in some way- not if you’re trying to bring some clarity to your own thoughts, much less others. You don’t have to land on a conclusion either. By all means chew over things, but once you try to settle, things fall apart if you haven’t done the harder work of replacing the framework of law with that of another conceptual structure.
You need to at least be asking “to what ends? What purpose is served by the rule?” Otherwise you’re stuck in things where half the time you end up arguing backwards in ways that put purpose serving rules, the maintenance of the rule with justifications ever further afield pulled in when the rule is questioned and edge cases reached. If you’re asking, essentially, “is the spirit of the rule still there?” You’ve got to stop and fill in what that spirit is or you or people that want to control you or have an agenda will sweep in with their own language and fill the void to their own ends.
Surprised they don't mention Google LLC v. Oracle America, Inc. Seems a bit myopic to condone the general legality while arguing "you can only use it how I like it".
It also doesn't talk about the far more interesting philosophical queston. Does what Blanchard did cover ALL implementations from Claude? What if anyone did exactly what he did, feed it the test cases and say "re-implement from scratch", ostensibly one would expect the results to be largely similar (technically under the right conditions deterministically similar)
could you then fork the project under your own name and a commercial license? when you use an LLM like this, to basically do what anyone else could ask it to do how do you attach any license to it? Is it first come first serve?
If an agent is acting mostly on its own it feels like if you found a copy of Harry Potter in the fictional library of Babel, you didn't write it, just found it amongst the infinite library, but if you found it first could you block everyone else that stumbles on a near-identical copy elsewhere in the library? or does each found copy represent a "Re-implementation" that could be individually copyrighted?
It should be noted that the Rust community is also guilty of something similar. That is, porting old GPL programs, typically written in C, to Rust and relicensing them as MIT.
Someone be brave, and do this to ZFS. Poke the Oracle bear!
Broadly speaking, the “freedom of users” is often protected by competition from competing alternatives. The GNU command line tools were replacements for system utilities. Linux was was a replacement for other Unix kernels. People chose to install them instead of proprietary alternatives. Was it due to ideology or lower cost or more features? All of the above. Different users have different motivations.
Copyleft could be seen as an attempt to give Free Software an edge in this competition for users, to counter the increased resources that proprietary systems can often draw on. I think success has been mixed. Sure, Linux won on the server. Open source won for libraries downloaded by language-specific package managers. But there’s a long tail of GPL apps that are not really all that appealing, compared to all the proprietary apps available from app stores.
But if reimplementing software is easy, there’s just going to be a lot more competition from both proprietary and open source software. Software that you can download for free that has better features and is more user-friendly is going to have an advantage.
With coding agents, it’s likely that you’ll be able to modify apps to your own needs more easily, too. Perhaps plugin systems and an AI that can write plugins for you will become the norm?
> Was it due to ideology or lower cost or more features?
It was due to access.
Why are people even having problems with sharing their changes to begin with? Just publishing it somewhere does not seem too expensive. The risk of accidentally including stuff that is not supposed to become public? Or are people regularly completely changing codebases and do not want to make the effort freely available, maybe especially to competitors? I would have assumed that the common case is adding a missing feature here, tweaking something there, if you turn the entire thing on its head, why not have your own alternative solution from scratch?
You can't put a copyright and MIT license on something you generated with AI. It is derived from the work of many unknown, uncredited authors.
Think about it; the license says that copies of the work must be reproduced with the copyright notice and licensing clauses intact. Why would anyone obey that, knowing it came from AI?
Countless instances of such licenses were ignored in the training data.
IMHO, the API and Test Suite, particularly the latter, define the contract of the functional definition of the software. It almost doesn't matter what that definition looks like so long as it conforms to the contract.
There was an issue where Google did something similar with the JVM, and ultimately it came down to whether or not Oracle owned the copyright to the header files containing the API. It went all the way to the US supreme court, and they ruled in Google's favour; finding that the API wasn't the implementation, and that the amount of shared code was so minimal as to be irrelevant.
They didn't anticipate that in less than half a decade we'd have technology that could _rapidly_ reimplement software given a strong functional definition and contract enforcing test suite.
"Antirez closes his careful legal analysis as though it settles the matter. Ronacher acknowledges that “there is an obvious moral question here, but that isn't necessarily what I'm interested in.” Both pieces treat legal permissibility as a proxy for social legitimacy. "
This whole article is just complaining that other people didn't have the discussion he wanted.
Ronacher even acknowledged that it's a different discussion, and not one they were trying to have at the moment.
If you want to have it, have it. Don't blast others for not having it for you.
Having this discussion involves blasting others for not considering it. Consider the rest of the paragraph you quoted:
> But law only says what conduct it will not prevent—it does not certify that conduct as right. Aggressive tax minimization that never crosses into illegality may still be widely regarded as antisocial. A pharmaceutical company that legally acquires a patent on a long-generic drug and raises the price a hundredfold has not done something legal and therefore fine. Legality is a necessary condition; it is not a sufficient one.
If the discussion inherently cannot be had without blasting innocent bystanders, I don't think it's a discussion worth having.
It might even be morally abhorrent to have such a discussion in the first place!
> If source code can now be generated from a specification, the specification is where the essential intellectual content of a GPL project resides. Blanchard's own claim—that he worked only from the test suite and API without reading the source—is, paradoxically, an argument for protecting that test suite and API specification under copyleft terms.
This is an interesting reversal in itself. If you make the specification protected under copyright, then the whole practice of clean room implementations is invalid.
Not a lawyer, but my understanding is: In theory, copyright only protects the creative expression of source code; this is the point of the "clean room" dance, that you're keeping only the functional behavior (not protected by copyright). Patents are, of course, an entirely different can of worms. So using an LLM to strip all of the "creative expression" out of source code but create the same functionality feels like it could be equivalent enough.
I like the article's point of legal vs. legitimate here, though; copyright is actually something of a strange animal to use to protect source code, it was just the most convenient pre-existing framework to shove it in.
> When GNU reimplemented the UNIX userspace, the vector ran from proprietary to free. Stallman was using the limits of copyright law to turn proprietary software into free software. […] The vector in the chardet case runs the other way.
That’s just your subjective opinion which many other people would disagree. I bet Armin Ronacher would agree that an MIT licensed library is even freer than an LGPL licensed library. To them, the vector is running from free to freer.
Why does anyone need his new library? They can do what he did and make their own.
I'm glad we can fork things at a point and thumb our noses at those who wish to cash in on other's work.
Why would I make my own? The new library is released under MIT license and faster than the old one.
If you decide to improve it in any way to fit your needs you can merely tell your own AI to re-implement it with your changes. Then it's proprietary to you.
I feel like the licenses that suffer the most isn't the GPL, but the ones like SSPL. If your code can be re-implemented easily and legally by AWS using an LLM, why risk publishing it?
It does feel like open source is about to change. My hunch is that commercial open source (beyond the consultation model) risks disappearing. Though I'd be happy to be proven wrong.
It's clear that we're entering a new era of copyright _expectations_ (whether we get new _legislation_ is different), but for now realise this: the people like me who like copyleft can do this too. We can take software we like, point an agent at it, and tell it to make a new version with the AGPL3.0-or-later badge on the front.
But the LLM contributions would likely be ruled public domain, so AGPL may not be enforceable on these.
The point of GPL is to restrict distribution. If there’s already an MIT version, it’s useless.
but the point of an EULA is to restrict distribution, so AGPL3 can help there.
I'm less concerned about AI eroding copyleft and more exited about AI eroding copy right.
I don't think this part is correct: "If you distribute modified code, or offer it as a networked service, you must make the source available under the same terms."
That's what something like AGPL does.
One of the things that irks me about this whole thing is, if it’s so clean room and distinct, why make the changes to the existing project? Why not make an entirely new library?
The answer to that, I think, is that the authors wanted to squat an existing successful project and gain a platform from it. Hence we have news cycle discussing it.
Nobody cares about a new library using AI, but squash an existing one with this stuff, and you get attention. It’s the reputation, the GitHub stars, whatever
I mean, Blanchard was the longtime maintainer of chardet already, and had wanted to relicense it for years. So I think that complicates your picture of "squatting an existing successful project".
Honestly it's a weird test case for this sort of thing. I don't think you'd see an equivalent in most open source projects.
I agree. But you can't copyright goodwill and reputation. Trademark does provide some protection there, right?
A lot of untagged IANAL takes here today.
Perhaps software patents may play an even bigger role in the future.
Or, hopefully, even less of a role.
Imagine if the author has his way, and when we have AI write software, it becomes legally under the license of some other sufficiently similar piece of software. Which may or may not be proprietary. "I see you have generated a todo app very similar to Todoist. So they now own it." That does not seem like a good path either for open source software or for opening up the benefits of AI generated software.
What if someone doesn't declare that it has been reimplemented using an LLM? Isn't it enough to simply declare that you have reimplemented the software without using an LLM? Good luck proving that in court...
One thing is certain, however: copyleft licenses will disappear: If I can't control the redistribution of my code (through a GPL or similar license), I choose to develop it in closed source.
Arguably, the GPL has always been the wrong choice if you want to authoritatively control redistribution.
If the model wasn't trained on copyleft, if he didn't use a copyleft test suite and if he wasn't the maintainer for years. Clearly the intent here is copyright infringement.
If you have software your testsuite should be your testsuite, you do dev with a testsuite and then mit without releasing one. Depending on the test-suite it may break clean room rules, especially for ttd codebases.
I think what is happening is the collapse of the “greater good”. Open source is dependent upon providing information for the greater good and general benefit of its readers. However now that no one is reading anything, its purpose is for the great good of the most clever or most convincing or richest harvester.
Perhaps we should finally admit that copyright has always been nonsense, and abolish this ridiculous measure once and for all
I think AI is very much eroding the legitimacy of copyright - at least to software, which is long been questioned since it's more like math than creative expression.
I think the industry will realize that it made a huge mistake by leaning on copyright for protection rather than on patents.
Probably a wiser approach is to consider different times require different measures (in general!).
I did not study in detail if copyright "has always been nonsense", but I do agree that nowadays some of the copyright regulations are nonsense (for example the very long duration of life + 70 years)
shall we now have to think about the tradeoffs in adopting
- proprietary
- free
- slop-licensed
software?
> Ronacher notes this as an irony and moves on. But the irony cuts deeper than he lets on. Next.js is MIT licensed. Cloudflare's vinext did not violate any license—it did exactly what Ronacher calls a contribution to the culture of openness, applied to a permissively licensed codebase. Vercel's reaction had nothing to do with license infringement; it was purely competitive and territorial. The implicit position is: reimplementing GPL software as MIT is a victory for sharing, but having our own MIT software reimplemented by a competitor is cause for outrage. This is what the claim that permissive licensing is “more share-friendly” than copyleft looks like in practice. The spirit of sharing, it turns out, runs in one direction only: outward from oneself.
This argument makes no sense. Are they arguing that because Vercel, specifically, had this attitude, this is an attitude necessitated by AI, reimplementation, and those who are in favor of it towards more permissive licenses? That certainly doesn't seem to be an accurate way to summarize what antirez or Ronacher believe. In fact, under the legal and ethical frameworks (respectively) that those two put forward, Vercel has no right to claim that position and no way to enforce it, so it seems very strange to me to even assert that this sort of thing would be the practical result of AI reimplementations. This seems to just be pointing towards the hypocrisy of one particular company, and assuming that this would be the inevitable universal, attitude, and result when there's no evidence to think so.
It's ironic, because antirez actually literally addresses this specific argument. They completely miss the fact that a lot of his blog post is not actually just about legal but also about ethical matters. Specifically, the idea he puts forward is that yes, corporations can do these kinds of rewrites now, but they always had the resources and manpower to do so anyway. What's different now is that individuals can do this kind of rewrites when they never have the ability to do so before, and the vector of such a rewrite can be from a permissive to copyleft or even from decompile the proprietary to permissive or copyleft. The fact that it hasn't been so far is a more a factor of the fact that most people really hate copyleft and find an annoying and it's been losing traction and developer mind share for decades, not that this tactic can't be used that way. I think that's actually one of the big points he's trying to make with his GNU comparison — not just that if it was legal for GNU to do it, then it's legal for you to do with AI, and not even just the fundamental libertarian ethical axiom (that I agree with for the most part) that it should remain legal to do such a rewrite in either direction because in terms of the fundamental axioms that we enforce with violence in our society, there should be a level playing field where we look at the action itself and not just whether we like or dislike the consequences, but specifically the fact that if GNU did it once with the ability to rewrite things, it can be done again, even in the same direction, it now even more easily using AI.
> They completely miss the fact that a lot of his blog post is not actually just about legal but also about ethical matters.
Honestly I was confused about the summarization of my blog post into just a legal matter as well. I hope my blog post will be able to flash at least a short time in the HN front page so that the actual arguments it contain will get a bit more exposure.
I'm failing to see what in the quoted text you took to be about AI rewrites specifically? It just reads as a slightly catty aside about the social reaction of rewrites in general (by implying the one example is generalizable.)
I think we're going one step too far even, AI itself is a gray area and how can they guarantee it was trained legally or if it's even legal what they're doing and how can they assert that the input training data didn't contain any copyrighted data.
Google already spent billions of dollars and decades of lawyer hours proving it out as fair use. The legal challenges we see now are the dying convulsions of an already broken system of publishers and IP hoarders using every resource at their disposal to manipulate authors and creators and the public into thinking that there's any legitimacy or value underlying modern copyright law.
AI will destroy the current paradigm, completely and utterly, and there's nothing they can do to stop it. It's unclear if they can even slow it, and that's a good thing.
We will be forced to legislate a modern, digital oriented copyright system that's fair and compatible with AI. If producing any software becomes a matter of asking a machine to produce it - if things like AI native operating systems come about, where apps and media are generated on demand, with protocols as backbone, and each device is just generating its own scaffolding around the protocols - then nearly none of modern licensing, copyright, software patents, or IP conventions make any sense whatsoever.
You can't have horse and buggy traffic conventions for airplanes. We're moving in to a whole new paradigm, and maybe we can get legislation that actually benefits society and individuals, instead of propping up massive corporations and making lawyers rich.
Google has cut out some very specific ruling that have nothing to do with modern AI. These systems are just a really slow/lossy git clone, current law has no trouble with it, it's broadly illegal.
If corporations are allowed to launder someone else work as their own people will simply stop working and just start endlessly remixing a la popular music.