The Corgi event doesn't seem particularly notable. There are similar features implemented in the most bog standard way that those features can be implemented using the pattern that AFAIK Github pioneered with a 'Danger Zone'. Both parties are using the same upstream components so it ends up looking the same.
I don't know when the extreme intellectual property viewpoint entered software engineering as a mainstream opinion because I have never before seen it expressed so strongly in this community (seeing as I wasn't around when Bill Gates famously asked for money first or whatever). To think that a past OpenOffice would have been considered unconscionably close to a copy of an old MS Office of the era twenty years ago.
In some way, The Corporations Won, because it turns out software engineers turned into IP maximalists. Thinking back to when I first installed Tux Kart decades ago I never could have imagined that we'd get to this stage. Really wild, man.
"I don't know when the extreme intellectual property viewpoint entered software engineering as a mainstream opinion because I have never before seen it expressed so strongly in this community"
It's not copyright maximalism, it's just bog-standard rationalization. I don't like what this company is doing, it looks like I can hit them with the "copyright" stick, so I will. One day later, I like playing abandonware games and that should be legal and copyright is stopping me so copyright bad, grrr argh.
At least, at the HN gestalt level. Individuals may say one or the other of those things from a principled perspective, but I perceive a lot of rationalization in these discussions overall.
There's not a lot people coming at this from any sort of principled position. I think one measure of that is that the modal principled position right now ought to be something fairly close to "I don't know". AI has kicked a lot of the foundation out from underneath copyright and I don't think anyone serious has more than a first draft of what the plan moving forward should be. Even if you can get two people to agree on the goals we should shoot for, which is already a tough ask even in a pre-AI era, getting them to agree on how to achieve those goals will be a long shot... and that's entirely separate from the question of whether the actions would in fact end up accomplishing the goal, which I don't trust anyone to have a good bead on right now.
Nominally, the principle of copyright has been to preserve creativity. Ten years ago we all had a reasonably similar idea what that meant, but we don't even have that now.
> At least, at the HN gestalt level. Individuals may say one or the other of those things from a principled perspective, but I perceive a lot of rationalization in these discussions overall.
I'd be careful attributing anything to the "HN gestalt"; it's a very, VERY wide range of individuals, with widely-ranging views. And I've been surprised, at times, when I've posted something that I thought would be widely disagreed with and downvoted, and yet my comment ended up with 5 net positive votes. Because the "gestalt", which I would call the consensus, on any given thread depends entirely on who feels invested enough in that topic to click on the thread and vote on it.
So on one thread you might find a lot of people holding position A, then on the other thread the vast majority is expressing position not-A, in direct contradiction. "Oh," you might conclude, "the HN gestalt is self-contradictory"... but if you were to actually dig into the comments and put together a spreadsheet of names and what they were advocating, you might find that most individuals were being consistent; it's just that there were largely different people posting on the two threads. (And some A advocates were posting their A advocacy on the second thread, but being drowned out by the majority of voices on that thread; while the first thread had a few not-A advocates, but not very many).
As for copyright, I've long felt that "death of the author + X years" was a bad system, and was worse as the value of X kept on being bumped up. I think it should simply be "X years", period, so it's predictable. For a reasonably large value of X, such as 50 years: authors who write a masterpiece in their 20's should still get to profit from its sales until they're 70+ years old. (And most authors don't just write one book and stop, so unless that hypothetical author is a one-and-done writer, he/she would still have many other books to profit from when that first book lapsed into the public domain).
But I haven't given much thought yet to what it looks like once AI use is common. (And if you think AI use is common now, just wait until open models start taking off in popularity, and AI use no longer requires a subscription fee. Might take a while for hardware to come down in price, so it might be 10 years instead of 5, but there's going to be a definite shift in lots and lots of ways once many more people can just pay a one-time hardware price rather than an ongoing subscription or per-token API price). So I can't really offer much else to the conversation than that.
> The Corporations Won, because it turns out software engineers turned into IP maximalists
Yeah, it's always bothered me that developers are eager to get legal to patent something they were involved in at Company™ put their name on patents so they can include them in their promotion docs.
It's like we're slowly defrauding ourselves and ensuring companies have the maximum legal standing against individual makers. Almost like a prisoner's dilemma where we're building the things that will/have be used against it us and those that come after in exchange for a little personal gain.
Corgi was spending 10K/yr for DocSend. They decided to build their own and the LLM took UI from PaperMark, an open-source alternative. I’ve implemented features that exist in open-source and LLM has never blatantly ripped off the UI.
Corgi has raised $106M but won’t pay for DocSend? Fine. Wants to build their own version based on Open-source? Fine. Keep it closed source and then try to monetize your version? That’s treacherous.
How is Microsoft Office vs Open Office a good parallel to this?
MSFT declared open source a "cancer" and "a threat to American Values" .. later, almost the entire Internet is run fundamentally on OSS.
What motivations might MSFT have had in 1998 ? Are there clear lessons from the extremes of the past that could be applied now?
Do authors have a right to LICENSE software they write? which ones, only Linus or Daniel Stenberg ? are there others? does a LICENSE mean anything ?
##-- related
From: "Eric S. Raymond" <esr@thyrsus.com>
To: wire-service@thyrsus.com
Subject: Microsoft and the Big Lie
Date: Fri, 1 Jun 2001 16:47:38 -0400
...
In the last three months, Jim Allchin and Craig Mundie and Steve Ballmer ... have described it (open source code ed.) as "un-American", "a destroyer", and "a cancer". They have deliberately confused the GPL with non-infectious open-source licenses, and they have deliberately confused active combination of code with passive aggregation of data.
OpenAI/et al. selling an IP laundering service under the name 'max subscription' may force the world to accept the perspective that Intellectual Property isn't a thing. The business model of extracting value from creators via rent seeking IP may not be viable in a world where LLMs can generate anything on demand. We might be transitioning to the Lockean view that for something to be ownable as property, it must be a scarce resource, and information is not a scarce resource.
From that property rights perspective, the property that's created when new information is created is not the information itself, rather, it's the act of creation (claim to authorship) that's the scarce resource.
I don't know what a world looks like where the only form of IP is non-transferable and owned by the original creator. Maybe that new form of IP creates less value over all, and maybe that's ok if the creator is getting 100% of the smaller pie instead of crumbs from media labels. Companies like Red Hat could be an example of a viable business model if IP laws follow the current winds.
Companies like Corgi will need to rely on internal talent to ensure that their product is better than what someone looking at their product can vibe code a copy of, which from my perspective as a consumer, sounds like a better route than Corgi relying on an internal legal team to send a cease and desist letter.
"Now software developers are feeling what authors and artist felt".
As an artist who got repeatedly told to stop making buggy whips and get into the absolutely tedious-sounding new field of "writing prompts" every time I expressed dismay and displeasure about image generation around here, every story about this sort of thing here is the sweetest schadenfreude I have tasted in my life.
Especially when the general feeling in the markets I work in is that AI images are kinda tacky and empty and nasty, and people would rather pay another human to realize their ideas than try to refine image generation prompts for a couple hours and get something vaguely okay that makes people go "ew, AI".
I can’t imagine software engineers caring about this at all. The only people that care about UI theft are C suite IP clutchers. SWEs generally love being able to use each other’s work. Copying UI by eye has been SOP since the dawn of computing.
Copyright doesn't cover instructions like recipes, protocols, or APIs; those require patents.
Not looking at the source code has been used to make nuisance copyright lawsuits less likely (e.g. Phoenix and AMI implementations of IBM's BIOS) but it's still easy to prevail when a new work is created by rewriting some else's source code. (https://en.wikipedia.org/wiki/UNIX_System_Laboratories,_Inc.....)
> Registered designs provide protection for the visual appearance of a product. In the context of GUIs, this means that the unique visual elements of your digital interfaces (i.e. the lines, contours and shapes) can be safeguarded against unauthorised use by competitors.
It looks like a registered design is similar to a design patent in the US. The automatic copyright is the same in the US as the UK, probably due to the Berne Convention.
Apple did file a claim against Samsung for multiple design patent infringements, including a design patent for using a grid of icons with rounded corners, but the US Supreme Court ruled that Apple would have to show financial loss from the infringement, which basically gives it no more power than a trademark. That was after a jury trial that heavily favored Apple, so it's not likely that anyone will ever get anything meaningful out of trying to enforce a design patent on a user interface alone.
Has anything like Apple's lawsuits against Microsoft or Samsung happened in the UK? Outside of purposeful counterfeiting, case law over trademark and design patent clams seems to be pretty weak. Then again, copyright is only powerful because of centuries of case law expanding its reach, and as written probably shouldn't cover fan fiction, shared universes, or possibly even translations. Intellectual property in general seems to be more about common law than the actual statues creating it.
Copyright doesn't cover the results of code, nor the methods used in the code, techniques and algorithms aren't covered by copyright. Period. Copyright applies to 'the work'. If you don't copy the source code, it's not covered.
Designs and strings are only sometimes covered under copyright.
The elements of a design that are ornamental, utilitarian, or a general look and feel are not covered under copyright but would be covered under a design patent if one exists.
Strings are only covered under copyright if they are a sufficiently original work of human expression. Simple informational messages generally wouldn't qualify.
This was copying many screens, composing layouts and many strings. not a "general look and feel" or one off utility string. Others found parts of the code copied (exact variable names). It would take a court to decide, but I think it's pretty clear cut copyright violation. They gave their agent had access to the EE code, and seemingly asked it to produce a copy. It's not a 1:1 copy, but there was a lot of copying. No clean-room attempt was even made.
If I take the first chapter of Harry Potter and write a new ending, it's still copyright violation. They took many parts.
If they did take sufficiently significant parts of copyrightable components, then yes, it would be a copyright violation. I don't think I saw overwhelming evidence of that in the OP, but certainly there is more to the story.
Whatever LLM they used copied the source code. It took their prompt and filled in the blanks of the spec by copying from the closest matching open source project. This is just what a next token predictor is going to do if you tell it to replicate software that's in its training set since that is the optimal way to predict the next tokens.
"still"? It never was. If you copy a (copyrighted) UI in bulk, that's a copyright violation just like copying code in bulk. The legal metric is generally "sufficient height of creation", the actual interpretation depends on where you are.
Copyrights on UIs are nebulous. You can't copyright the functional aspects of a UI, that's the domain of patents, yet the functional aspects are likely the parts you are most keen on protecting. Also you need to prove that they copied you and didn't come up with it independently.
I looked at those screen shots. The language is similar but it’s not verbatim. The is itself is just the usage of the same framework. Business logic can’t be copyrighted. Is it too close? I agree. Does it cross the legal test. I wouldn’t waste my money on a court battle.
I wasn't talking about a specific case, and "UI" is more than a single page. To be specific, I'm referring to the sum of text (incl. translations), graphics and layout on the entire product/application/etc.
I agree the screenshots in the specific linked case - if that's all there is - are nowhere near enough.
It depends also what you mean with copy. The code has copyright but the rendered pixels don’t (other than if something like an image was created pixel by pixel). So if the code is different but the output looks the same it’s not copyright infringement (can still be trademark).
I think that’s dependent on jurisdiction. This comes up with fonts where what you said applies in the US but in Europe, the actual visual result of the software also counts and is protected (afaik).
I don’t know any jurisdiction that gives copyright to the rendered output of html/css. There can be other IP (patents, pattern, trademark but doubt they apply for a normal saas ui)
I don’t really see why it’s hard to believe that the design of a website wouldn’t be protected. Obviously a digital painting is protected, so hand writing svg code to generate an image would also result in an image that’s protected by copyright. And why would a html website not be protected?
The law doesn’t really care about how you create the work as long as it is an expression of your own creativity.
> Es ist allgemein anerkannt, dass der Gestaltung von Webseiten unabhängig von der Digitalisierung ihres Inhalts ein Urheberrechtsschutz zukommen kann, sofern die Gestaltung die gemäß § 2 Abs. 2 UrhG erforderliche Schöpfungshöhe erreicht (OLG Frankfurt a.a.O.; OLG Hamm, MMR 2005, 106; OLG Düsseldorf MMR 1999, 729; LG München I MMR 2005, 267). Dies kommt aber grundsätzlich nur dann in Betracht, wenn die Gestaltung der Webseiten über das hinausgeht, was bei ordnungsgemäßer Erstellung eines Webauftritts im Internet handwerklich zu leisten ist.
If you're worried about infringement, register your work with the US copyright office. You can only get monetary and statutory damages if the work was registered before infringement, otherwise you can only get an injunction. But you can't even file a claim in court to request an injunction without first registering the work. Basically, while copyright nominally attaches at creation, without a certificate you can't press any rights in court.
You don't need to register each release, so long as a material portion of the registered work exists in subsequent derivative works.
Without a registration threats of a copyright dispute are mostly noise to someone savvy enough to know how the game is played. If they think you'll persist they can just replace the infringing work or cease distribution, which is a hassle but not a significant deterrence for bad faith actors.
No, that was never a defense. Replication of the work artifacts without literal copy/paste can still be a violation. It depends on just how similar the works are as demonstrated to a civil court jury.
At the same time complete recreations are not copyright violations. For example ZeldaClassic.com is not a violation of copyright against the original game. The final products appear almost like a clone, but the underlying work is completely different. Copyright law protects work artifacts, not ideas or assembled constructions.
Software copyrights are among humanity's worst inventions. We as a species are no better off because of it, and neither are the small creators that copyrights are supposed to protect. Software copyrights only exist to protect a renter model from big corporations.
There's an argument to be made for patent protections, but many of those are questionable considering the number of trivial software-related patents (there must be a patent somewhere for replying to an online conversation through an edit box and an "add comment" button).
I don't know if LLMs can somehow help the situation. I hope they can expose the ridiculousness of software copyrights but I won't be holding my breath.
> I hope they can expose the ridiculousness of software copyrights but I won't be holding my breath.
I think it has already, but it's a hard pill for many to swallow. While I haven't actually counted, the conversations around the effectiveness of copyright/IP seem to be regular conversations now, both here and IRL.
Software is a particularly egregious case, but the problem is more general. Copyright itself is just a bad move all around.
Copyright demands that everyone pretend the value of someone's work is the product of that work, not the labor. Therefore, we should not expect people to earn wages for labor; and we should instead expect people to earn royalties from their "works" (the countable commodity). Absurd.
Copyright grants "artists" (in the broadest sense of the word) a monopoly over their "work", again the imagined product of their labor. In practice, this actually means a monopoly on the labor itself, because all art is derivative work, and the derivation of work is the specific thing that copyright monopolizes. Twice absurd.
LLMs, in the best case, are calling that bluff. The problem is that they are calling it poorly, and the bluff itself is incoherent to begin with. Even worse is that LLMs can be monopolized as copyrighted "works", which is a clear abuse of the system.
We should get rid of copyright and patents. Dismantle all the moats and publishing houses (including social media). Liberate derivative work. Value labor directly.
"Copyright demands that everyone pretend the value of someone's work is the product of that work, not the labor."
Isn't it?
If you spend ten years writing the Great American Novel, and I spend ten years writing "All work and no play makes Jack a dull boy" over and over again, have I created as much value as you have?
If I would have spent ten years writing the Great American Novel, but spent all my time working elsewhere so I could afford to live, then yes.
Why in the hell would anyone spend ten years writing "All work and no play makes Jack a dull boy" over and over again? Because they have no intrinsic meaning to provide, or because their life has no meaning to reflect?
We may as well set aside this argument anyway, because it actually isn't relevant. If I accept your premise that the product of work is the only value in labor, then why are there specific categories of product that I can value, and others that I cannot? That's the situation copyright has put us in: if I create the right kind of work, but it's derivative, then that's a violation of someone's copyright. If I create derivative work, but it's the wrong kind, then I can't copyright it. The only kind of work that I can profit from freely is "original", which is a false premise to begin with.
So what is the alternative? Speculation. In a society without copyright, labor must be funded somehow. Rather than promise a pot of gold at the end of a rainbow, we would simply fund artists who we believe will create valuable products. We already do this to a moderately significant extent: everyone knows about Patreon and OnlyFans. Most successful creators rely on ad revenue instead of royalties. The problem with this model is that it must compete with copyright holders, who get to monopolize entire swaths of derivative work, and leverage the guarantee of their already-performed work as much easier to sell than speculative investment. Get rid of copyright, and the market simply becomes fair.
Of course no sane person would spend ten years writing "All work and no play makes Jack a dull boy" over and over again". That wasn't the point.
The labor theory of value says that the value of something is determined by the amount of labor that went into creating it. I'd say it's not really so much a theory as a definition of the word "value" -- and I suggest that it's not a particularly useful definition.
If I go to the widget store to buy a widget, the price I'm willing to pay depends on what it's worth to me. The amount of work that went into producing this widget or that widget doesn't affect my decision.
> Everyone pretend the value of someone's work is the product of that work, not the labor.
Is it not? If I spend 10 years writing the greatest novel of all time, and you, a publishing company, make copies and sell 10 million copies, I feel entitled to some recompense.
My labor has value to me, but only the product of that labor has value to anyone else.
You feel entitled to recompense from the publisher, because they are profiting from your labor. The problem I brought up is still there, just in the hands of the publisher, and not the writer. Now the publisher gets to profit from work they did not do, and you are lucky to get a good cut.
Your labor has value to everyone else, even if that value is speculative. If we don't have a mechanism to commoditize the ends of labor, then we can just speculate instead. Speculating the value of labor is more uncertain than valuing copyrighted works, which means that the business of labor speculation cannot compete with the business of copyright valuation. At the same time, copyright is a lie: the "product of a work" is a totally arbitrary imagined boundary that can't always be meaningfully drawn in the first place; meaning that entire categories of work are impossible to copyright at all. Removing that lie would put everyone on a level playing field, where all labor is valuable, and all valuation is fair.
In effect the source code is being copied by the LLM. This is what it's designed to do. LLMs are a lossy statistical compression of their training data.
If you give it a prompt telling it to replicate a product that's in its training set then its optimal next token prediction output is going to be to a lossy copy of that product's source code.
There are no novel UIs, so copying UIs is okay, and necessary. As for source code, I'm a stickler for the license. The modern set of licenses cover any scenario I can think of, relatively fairly. AI is merely a tool, so the craftsman still owns the output. If the output violates a license, then the craftsman should be held to account.
The look and feel of an interface could be covered under a design patent, which would be a different question than copyright, and a much higher bar to enforce (particularly, protection is far from automatic).
You’re talking about enforcement which is a fair point of view, but I’m taking about likelihood of getting sued and having to settle in order to avoid the risk of losing a case in court. Which is the more common case for situations where one company alleges the other stole their IP.
The replication/copying has always been there in one form or another. The bar has traditionally been higher for reputation and monetary risks.
Lately the legal bar is the one that going down, ease of replication makes it even more tempting and when big players are doing it at scale (bots) then it validates the strategy in one way or another.
If anything, there have to be downstream consequences of this with time, libraries to pollute the front end code for LLMs are most likely going to get popular and probably one way to make it harder for your IP to be replicated.
Think it'll be hard to define in any sufficiently specific manner legally because UI/UX/source/functionality aren't entirely separate, but ethically I reckon it comes down to what one means by copy UI.
e.g. If you're creating an uptime dashboard...they all kinda look the same anyway and there aren't that many ways to do it so that seems OK. If it's copying an comprehensive UI with layout and flow between the various pages etc then you're getting a bit closer to theft.
It depends. Is the text auto-generated from a framework? Is it creative? Is it worthy of copyright? It seems more instructional.
I say let them sue for copyright infringement for the text. Let them sue for breaking a license. Let's see how it works out. Doesn't impact me and if it costs the rich money, good. Let them suffer.
So to be clear the answer is emphatically "no". If you copy everything else, the defense that the source code is technically different will not save you.
copying other business pixel-to-pixel and direct claims agains competitor is just distasteful. add to this recent YC "AI that MITM API and re-implments anything automatically" is very bad image to YC.
with such bad behavior from SWE community, you just got to lock down your app behind certificate pinning, hardware attestation, gRPC/protobufs, and internal data only. no more "free open web in browsers" when you get gents like this stealing other peoples efforts.
I made a leap there without explaining. You can copyright code the same way you copyright a book. But in doing so you protect the text but not inherently the IP. Copyrighted code has a whole "originality" argument in it's defensibility. Patents do too, but they would never patent the code itself. If you want to protect yourself... it's belt and suspenders.
...hopefully as other comments said, that LLMs make us abandon the confusing idea of "intellectual property" so we don't have to ask questions like this or get in to torturous questions if this or that thing is "infringing" or not
There's a couple of related issues being conflated here, and I'm not sure which one to bring up, mainly because I'm not sure in what direction the copying would be ruled to have gone. So I'll just mention all the cases.
The first thing to note is that nonliteral copying can still be infringing. Actually, among copyright cases that actually go to trial, most of them are not bit-exact matches ("striking similarity" in legalese). The lower standard those cases would have to meet is substantial similarity, which requires proving both access and similarity. In other words, in order for you to produce a copy[0], you have to have both seen the original and produced something that is close enough if you squint at it.
So let's say Papermark is the original and Corgi Dataroom copied it. If this actually went to trial, a significant amount of discovery would be spent harvesting all the e-mails and messages Corgi's development team sent to one another. Any evidence of knowledge or access to Papermark would probably be enough to prove a copyright violation.
You mentioned LLMs, and some of the tweets here also mention them. I have no clue if either product used an LLM, but it's important to note that in the US, anything written by an LLM does not accrue copyright protection. So, if Papermark was LLM-authored, as a threshold matter, they would have to register[2] a very specific copyright that neatly carves out the LLM-authored bits. The judge would then only consider the parts of the code with verifiable human authorship, which would severely weaken Papermark's case.
In the reverse case - i.e. Corgi Dataroom is LLM-authored - then Papermark's case becomes stronger. Note how I didn't say "AI slop is public domain" last paragraph, because it isn't. There are still unsettled legal questions as to whether or not training on copyrighted works is legal and if using an LLM trained on that work constitutes access to it. Furthermore, LLMs can have search tools that would give them access to code not within the training set, which would also be a more straightforward copyright violation. So if it turns out Dataroom's developers are all Claude fans, and Claude is copying Papermark code, then it's just a normal license violation. LLMs do not launder copyright.
We can also consider the case where BOTH tools are LLM-authored. In this case, there might just not be a copyright case at all. Technically speaking, there would be some third class of plaintiffs who have been infringed, but they would have to choose to sue. It is a long-standing principle in law that you are not allowed to sue for other people's harms[1]. So in this case, nobody would have a case.
> Now software developers are feeling what authors and artist felt
It was specifically the FOSS community that sounded the alarm about training data theft first, because the FOSS community correctly understood coding agents to be an attack on copylefts[3] (albeit through the incorrect belief that LLMs could launder away copyright interest, as opposed to it just being told to make a noninfringing substitute).
[0] I am skipping over notions of fair use and derivative works as they would complicate the analysis and do not apply here.
[1] In general, the operating principle of American courts is "fuck around and find out", and this implies that the court is only allowed to find out once someone has fucked around. Otherwise, the courts could just sue themselves to rule on whatever the hell they want. Isn't adversarial common law GREAT!?
[2] Yes, copyright registration is mandatory in the US, otherwise you can't sue, which is the whole point of copyright. The Berne Convention only half-applies here.
[3] Clauses in licenses that require modifications to the work to be provided under the same license terms. Creative Commons calls these Share-Alike licenses.
> So if it turns out Dataroom's developers are all Claude fans, and Claude is copying Papermark code, then it's just a normal license violation. LLMs do not launder copyright.
That's an easy case, but a more interesting question to me is if there wasn't any direct source code access (similar to an open source example I can't remember the name of atm). If you give an LLM requirements matching the copyrighted products capabilities, and it doesn't have Internet access, and it spits out different code that accomplishes the same stuff papermarks product does, do you believe courts would allow that, and copyrighted could be laundered legally?
If, in fact, you copy computer software by spec and never had access to the code -- that's a clean room reimplementation and doesn't violate copyright. Copyright covers the human expression of software -- the text of the code itself. It can also cover non-code assets included in your software like graphics. Copyright does not cover the ideas, goals, designs, concepts, principles, etc of your software.
Did you agree to terms of use? Did you have to click a check box that you agree to terms of use before seeing or having access to the items you copied? Click wrap. If in the contract that you agreed to there is language that you agreed to not copy the work, then you likely are in breach of contract. If it is publicly available knowledge probably not breach of contract. I’m not a lawyer of course.
There is a lot of copying that isn't protected by copyright. It is possible to include what can't be copied that might not fall under copyright in the terms of service contract. Many people not being able to use copyright to prevent copying instead successfully sued based on breach of contract.
Companies should understand that they can protect their IP this way.
The Corgi event doesn't seem particularly notable. There are similar features implemented in the most bog standard way that those features can be implemented using the pattern that AFAIK Github pioneered with a 'Danger Zone'. Both parties are using the same upstream components so it ends up looking the same.
I don't know when the extreme intellectual property viewpoint entered software engineering as a mainstream opinion because I have never before seen it expressed so strongly in this community (seeing as I wasn't around when Bill Gates famously asked for money first or whatever). To think that a past OpenOffice would have been considered unconscionably close to a copy of an old MS Office of the era twenty years ago.
In some way, The Corporations Won, because it turns out software engineers turned into IP maximalists. Thinking back to when I first installed Tux Kart decades ago I never could have imagined that we'd get to this stage. Really wild, man.
"I don't know when the extreme intellectual property viewpoint entered software engineering as a mainstream opinion because I have never before seen it expressed so strongly in this community"
It's not copyright maximalism, it's just bog-standard rationalization. I don't like what this company is doing, it looks like I can hit them with the "copyright" stick, so I will. One day later, I like playing abandonware games and that should be legal and copyright is stopping me so copyright bad, grrr argh.
At least, at the HN gestalt level. Individuals may say one or the other of those things from a principled perspective, but I perceive a lot of rationalization in these discussions overall.
There's not a lot people coming at this from any sort of principled position. I think one measure of that is that the modal principled position right now ought to be something fairly close to "I don't know". AI has kicked a lot of the foundation out from underneath copyright and I don't think anyone serious has more than a first draft of what the plan moving forward should be. Even if you can get two people to agree on the goals we should shoot for, which is already a tough ask even in a pre-AI era, getting them to agree on how to achieve those goals will be a long shot... and that's entirely separate from the question of whether the actions would in fact end up accomplishing the goal, which I don't trust anyone to have a good bead on right now.
Nominally, the principle of copyright has been to preserve creativity. Ten years ago we all had a reasonably similar idea what that meant, but we don't even have that now.
> At least, at the HN gestalt level. Individuals may say one or the other of those things from a principled perspective, but I perceive a lot of rationalization in these discussions overall.
I'd be careful attributing anything to the "HN gestalt"; it's a very, VERY wide range of individuals, with widely-ranging views. And I've been surprised, at times, when I've posted something that I thought would be widely disagreed with and downvoted, and yet my comment ended up with 5 net positive votes. Because the "gestalt", which I would call the consensus, on any given thread depends entirely on who feels invested enough in that topic to click on the thread and vote on it.
So on one thread you might find a lot of people holding position A, then on the other thread the vast majority is expressing position not-A, in direct contradiction. "Oh," you might conclude, "the HN gestalt is self-contradictory"... but if you were to actually dig into the comments and put together a spreadsheet of names and what they were advocating, you might find that most individuals were being consistent; it's just that there were largely different people posting on the two threads. (And some A advocates were posting their A advocacy on the second thread, but being drowned out by the majority of voices on that thread; while the first thread had a few not-A advocates, but not very many).
As for copyright, I've long felt that "death of the author + X years" was a bad system, and was worse as the value of X kept on being bumped up. I think it should simply be "X years", period, so it's predictable. For a reasonably large value of X, such as 50 years: authors who write a masterpiece in their 20's should still get to profit from its sales until they're 70+ years old. (And most authors don't just write one book and stop, so unless that hypothetical author is a one-and-done writer, he/she would still have many other books to profit from when that first book lapsed into the public domain).
But I haven't given much thought yet to what it looks like once AI use is common. (And if you think AI use is common now, just wait until open models start taking off in popularity, and AI use no longer requires a subscription fee. Might take a while for hardware to come down in price, so it might be 10 years instead of 5, but there's going to be a definite shift in lots and lots of ways once many more people can just pay a one-time hardware price rather than an ongoing subscription or per-token API price). So I can't really offer much else to the conversation than that.
Anyone who refers to "the HN gestalt" already knows all that or they wouldn't have phrased it that way.
It isn't just me but it's a short list: https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...
> Both parties are using the same upstream components so it ends up looking the same.
I had a look at the 4 screenshots in the post, and definitively it's not just using the "same upstream components", it's a verbatim copy.
Don't know about the rest of the app and the actual code, but I won't be surprised to find out it's basically the same.
> The Corporations Won, because it turns out software engineers turned into IP maximalists
Yeah, it's always bothered me that developers are eager to get legal to patent something they were involved in at Company™ put their name on patents so they can include them in their promotion docs.
It's like we're slowly defrauding ourselves and ensuring companies have the maximum legal standing against individual makers. Almost like a prisoner's dilemma where we're building the things that will/have be used against it us and those that come after in exchange for a little personal gain.
Money always sprouts double standards out of everyone and anyone's ass
Did you even look into the situation?
Corgi was spending 10K/yr for DocSend. They decided to build their own and the LLM took UI from PaperMark, an open-source alternative. I’ve implemented features that exist in open-source and LLM has never blatantly ripped off the UI.
Corgi has raised $106M but won’t pay for DocSend? Fine. Wants to build their own version based on Open-source? Fine. Keep it closed source and then try to monetize your version? That’s treacherous.
How is Microsoft Office vs Open Office a good parallel to this?
> I don't know
it is abundantly clear from the post, agree
> Bill Gates
MSFT declared open source a "cancer" and "a threat to American Values" .. later, almost the entire Internet is run fundamentally on OSS.
What motivations might MSFT have had in 1998 ? Are there clear lessons from the extremes of the past that could be applied now?
Do authors have a right to LICENSE software they write? which ones, only Linus or Daniel Stenberg ? are there others? does a LICENSE mean anything ?
##-- related
From: "Eric S. Raymond" <esr@thyrsus.com> To: wire-service@thyrsus.com Subject: Microsoft and the Big Lie Date: Fri, 1 Jun 2001 16:47:38 -0400 ...
In the last three months, Jim Allchin and Craig Mundie and Steve Ballmer ... have described it (open source code ed.) as "un-American", "a destroyer", and "a cancer". They have deliberately confused the GPL with non-infectious open-source licenses, and they have deliberately confused active combination of code with passive aggregation of data.
I assume this is the Bill Gates reference:https://en.wikipedia.org/wiki/File:Bill_Gates_Letter_to_Hobb...
OpenAI/et al. selling an IP laundering service under the name 'max subscription' may force the world to accept the perspective that Intellectual Property isn't a thing. The business model of extracting value from creators via rent seeking IP may not be viable in a world where LLMs can generate anything on demand. We might be transitioning to the Lockean view that for something to be ownable as property, it must be a scarce resource, and information is not a scarce resource.
From that property rights perspective, the property that's created when new information is created is not the information itself, rather, it's the act of creation (claim to authorship) that's the scarce resource.
I don't know what a world looks like where the only form of IP is non-transferable and owned by the original creator. Maybe that new form of IP creates less value over all, and maybe that's ok if the creator is getting 100% of the smaller pie instead of crumbs from media labels. Companies like Red Hat could be an example of a viable business model if IP laws follow the current winds.
Companies like Corgi will need to rely on internal talent to ensure that their product is better than what someone looking at their product can vibe code a copy of, which from my perspective as a consumer, sounds like a better route than Corgi relying on an internal legal team to send a cease and desist letter.
> information is not a scarce resource
The powerful/corrupt will maintain their information asymmetry; software, tourism, self help etc will be devalued.
"Now software developers are feeling what authors and artist felt".
As an artist who got repeatedly told to stop making buggy whips and get into the absolutely tedious-sounding new field of "writing prompts" every time I expressed dismay and displeasure about image generation around here, every story about this sort of thing here is the sweetest schadenfreude I have tasted in my life.
Especially when the general feeling in the markets I work in is that AI images are kinda tacky and empty and nasty, and people would rather pay another human to realize their ideas than try to refine image generation prompts for a couple hours and get something vaguely okay that makes people go "ew, AI".
I can’t imagine software engineers caring about this at all. The only people that care about UI theft are C suite IP clutchers. SWEs generally love being able to use each other’s work. Copying UI by eye has been SOP since the dawn of computing.
> Copying UI by eye has been SOP since the dawn of computing.
Case in point, the very GUI operating systems we use today are arguably a "stolen" product from xerox parc.
As someone who has used multiple vibe coded internal tools: you will care when you use these tools and encounter strange bugs and missing features.
The human touch is visible in the way your features work just like in vibe coded art and games it lacks intention.
Copyright doesn't cover instructions like recipes, protocols, or APIs; those require patents.
Not looking at the source code has been used to make nuisance copyright lawsuits less likely (e.g. Phoenix and AMI implementations of IBM's BIOS) but it's still easy to prevail when a new work is created by rewriting some else's source code. (https://en.wikipedia.org/wiki/UNIX_System_Laboratories,_Inc.....)
Neither copyright nor patent cover a user interface (https://en.wikipedia.org/wiki/Apple_Computer,_Inc._v._Micros....), so that can legally be copied outright.
> Neither copyright nor patent cover a user interface
*us only
> You automatically get copyright protection when you create: ... original non-literary written work, such as software, web content and databases
https://www.gov.uk/copyright
> Registered designs provide protection for the visual appearance of a product. In the context of GUIs, this means that the unique visual elements of your digital interfaces (i.e. the lines, contours and shapes) can be safeguarded against unauthorised use by competitors.
https://cms.law/en/media/local/cms-cmno/files/protection-for...
It looks like a registered design is similar to a design patent in the US. The automatic copyright is the same in the US as the UK, probably due to the Berne Convention.
Apple did file a claim against Samsung for multiple design patent infringements, including a design patent for using a grid of icons with rounded corners, but the US Supreme Court ruled that Apple would have to show financial loss from the infringement, which basically gives it no more power than a trademark. That was after a jury trial that heavily favored Apple, so it's not likely that anyone will ever get anything meaningful out of trying to enforce a design patent on a user interface alone.
Has anything like Apple's lawsuits against Microsoft or Samsung happened in the UK? Outside of purposeful counterfeiting, case law over trademark and design patent clams seems to be pretty weak. Then again, copyright is only powerful because of centuries of case law expanding its reach, and as written probably shouldn't cover fan fiction, shared universes, or possibly even translations. Intellectual property in general seems to be more about common law than the actual statues creating it.
Copyright doesn't cover the results of code, nor the methods used in the code, techniques and algorithms aren't covered by copyright. Period. Copyright applies to 'the work'. If you don't copy the source code, it's not covered.
They copied strings, and visual design. Both of which are work, both of which are covered by copyright.
Designs and strings are only sometimes covered under copyright.
The elements of a design that are ornamental, utilitarian, or a general look and feel are not covered under copyright but would be covered under a design patent if one exists.
Strings are only covered under copyright if they are a sufficiently original work of human expression. Simple informational messages generally wouldn't qualify.
This was copying many screens, composing layouts and many strings. not a "general look and feel" or one off utility string. Others found parts of the code copied (exact variable names). It would take a court to decide, but I think it's pretty clear cut copyright violation. They gave their agent had access to the EE code, and seemingly asked it to produce a copy. It's not a 1:1 copy, but there was a lot of copying. No clean-room attempt was even made.
If I take the first chapter of Harry Potter and write a new ending, it's still copyright violation. They took many parts.
If they did take sufficiently significant parts of copyrightable components, then yes, it would be a copyright violation. I don't think I saw overwhelming evidence of that in the OP, but certainly there is more to the story.
Whatever LLM they used copied the source code. It took their prompt and filled in the blanks of the spec by copying from the closest matching open source project. This is just what a next token predictor is going to do if you tell it to replicate software that's in its training set since that is the optimal way to predict the next tokens.
"still"? It never was. If you copy a (copyrighted) UI in bulk, that's a copyright violation just like copying code in bulk. The legal metric is generally "sufficient height of creation", the actual interpretation depends on where you are.
Copyrights on UIs are nebulous. You can't copyright the functional aspects of a UI, that's the domain of patents, yet the functional aspects are likely the parts you are most keen on protecting. Also you need to prove that they copied you and didn't come up with it independently.
I looked at those screen shots. The language is similar but it’s not verbatim. The is itself is just the usage of the same framework. Business logic can’t be copyrighted. Is it too close? I agree. Does it cross the legal test. I wouldn’t waste my money on a court battle.
I wasn't talking about a specific case, and "UI" is more than a single page. To be specific, I'm referring to the sum of text (incl. translations), graphics and layout on the entire product/application/etc.
I agree the screenshots in the specific linked case - if that's all there is - are nowhere near enough.
It depends also what you mean with copy. The code has copyright but the rendered pixels don’t (other than if something like an image was created pixel by pixel). So if the code is different but the output looks the same it’s not copyright infringement (can still be trademark).
I think that’s dependent on jurisdiction. This comes up with fonts where what you said applies in the US but in Europe, the actual visual result of the software also counts and is protected (afaik).
I don’t know any jurisdiction that gives copyright to the rendered output of html/css. There can be other IP (patents, pattern, trademark but doubt they apply for a normal saas ui)
Germany? The creative work is protected, and the design can be a creative work. How you implement it doesn’t really matter.
Which paragraph? I doubt it really covers the final outputted design (unless its specifically some art project) of general ui designs.
paragraph 2
I don’t really see why it’s hard to believe that the design of a website wouldn’t be protected. Obviously a digital painting is protected, so hand writing svg code to generate an image would also result in an image that’s protected by copyright. And why would a html website not be protected?
The law doesn’t really care about how you create the work as long as it is an expression of your own creativity.
https://dejure.org/dienste/vernetzung/rechtsprechung?Text=2%...
> Es ist allgemein anerkannt, dass der Gestaltung von Webseiten unabhängig von der Digitalisierung ihres Inhalts ein Urheberrechtsschutz zukommen kann, sofern die Gestaltung die gemäß § 2 Abs. 2 UrhG erforderliche Schöpfungshöhe erreicht (OLG Frankfurt a.a.O.; OLG Hamm, MMR 2005, 106; OLG Düsseldorf MMR 1999, 729; LG München I MMR 2005, 267). Dies kommt aber grundsätzlich nur dann in Betracht, wenn die Gestaltung der Webseiten über das hinausgeht, was bei ordnungsgemäßer Erstellung eines Webauftritts im Internet handwerklich zu leisten ist.
If you're worried about infringement, register your work with the US copyright office. You can only get monetary and statutory damages if the work was registered before infringement, otherwise you can only get an injunction. But you can't even file a claim in court to request an injunction without first registering the work. Basically, while copyright nominally attaches at creation, without a certificate you can't press any rights in court.
You don't need to register each release, so long as a material portion of the registered work exists in subsequent derivative works.
Without a registration threats of a copyright dispute are mostly noise to someone savvy enough to know how the game is played. If they think you'll persist they can just replace the infringing work or cease distribution, which is a hassle but not a significant deterrence for bad faith actors.
No, that was never a defense. Replication of the work artifacts without literal copy/paste can still be a violation. It depends on just how similar the works are as demonstrated to a civil court jury.
At the same time complete recreations are not copyright violations. For example ZeldaClassic.com is not a violation of copyright against the original game. The final products appear almost like a clone, but the underlying work is completely different. Copyright law protects work artifacts, not ideas or assembled constructions.
Software copyrights are among humanity's worst inventions. We as a species are no better off because of it, and neither are the small creators that copyrights are supposed to protect. Software copyrights only exist to protect a renter model from big corporations.
There's an argument to be made for patent protections, but many of those are questionable considering the number of trivial software-related patents (there must be a patent somewhere for replying to an online conversation through an edit box and an "add comment" button).
I don't know if LLMs can somehow help the situation. I hope they can expose the ridiculousness of software copyrights but I won't be holding my breath.
> I hope they can expose the ridiculousness of software copyrights but I won't be holding my breath.
I think it has already, but it's a hard pill for many to swallow. While I haven't actually counted, the conversations around the effectiveness of copyright/IP seem to be regular conversations now, both here and IRL.
Software is a particularly egregious case, but the problem is more general. Copyright itself is just a bad move all around.
Copyright demands that everyone pretend the value of someone's work is the product of that work, not the labor. Therefore, we should not expect people to earn wages for labor; and we should instead expect people to earn royalties from their "works" (the countable commodity). Absurd.
Copyright grants "artists" (in the broadest sense of the word) a monopoly over their "work", again the imagined product of their labor. In practice, this actually means a monopoly on the labor itself, because all art is derivative work, and the derivation of work is the specific thing that copyright monopolizes. Twice absurd.
LLMs, in the best case, are calling that bluff. The problem is that they are calling it poorly, and the bluff itself is incoherent to begin with. Even worse is that LLMs can be monopolized as copyrighted "works", which is a clear abuse of the system.
We should get rid of copyright and patents. Dismantle all the moats and publishing houses (including social media). Liberate derivative work. Value labor directly.
"Copyright demands that everyone pretend the value of someone's work is the product of that work, not the labor."
Isn't it?
If you spend ten years writing the Great American Novel, and I spend ten years writing "All work and no play makes Jack a dull boy" over and over again, have I created as much value as you have?
https://en.wikipedia.org/wiki/Labor_theory_of_value#Critique...
If I would have spent ten years writing the Great American Novel, but spent all my time working elsewhere so I could afford to live, then yes.
Why in the hell would anyone spend ten years writing "All work and no play makes Jack a dull boy" over and over again? Because they have no intrinsic meaning to provide, or because their life has no meaning to reflect?
We may as well set aside this argument anyway, because it actually isn't relevant. If I accept your premise that the product of work is the only value in labor, then why are there specific categories of product that I can value, and others that I cannot? That's the situation copyright has put us in: if I create the right kind of work, but it's derivative, then that's a violation of someone's copyright. If I create derivative work, but it's the wrong kind, then I can't copyright it. The only kind of work that I can profit from freely is "original", which is a false premise to begin with.
So what is the alternative? Speculation. In a society without copyright, labor must be funded somehow. Rather than promise a pot of gold at the end of a rainbow, we would simply fund artists who we believe will create valuable products. We already do this to a moderately significant extent: everyone knows about Patreon and OnlyFans. Most successful creators rely on ad revenue instead of royalties. The problem with this model is that it must compete with copyright holders, who get to monopolize entire swaths of derivative work, and leverage the guarantee of their already-performed work as much easier to sell than speculative investment. Get rid of copyright, and the market simply becomes fair.
Of course no sane person would spend ten years writing "All work and no play makes Jack a dull boy" over and over again". That wasn't the point.
The labor theory of value says that the value of something is determined by the amount of labor that went into creating it. I'd say it's not really so much a theory as a definition of the word "value" -- and I suggest that it's not a particularly useful definition.
If I go to the widget store to buy a widget, the price I'm willing to pay depends on what it's worth to me. The amount of work that went into producing this widget or that widget doesn't affect my decision.
Another article you might (or might not) want to read: https://en.wikipedia.org/wiki/Criticisms_of_the_labour_theor...
"Get rid of copyright, and the market simply becomes fair."
I think the word "simply" is doing a whole lot of work there.
> Everyone pretend the value of someone's work is the product of that work, not the labor.
Is it not? If I spend 10 years writing the greatest novel of all time, and you, a publishing company, make copies and sell 10 million copies, I feel entitled to some recompense.
My labor has value to me, but only the product of that labor has value to anyone else.
The goal is the product, but what is being exchanged is production, ie labor for a certificate of your own past production, aka money.
You feel entitled to recompense from the publisher, because they are profiting from your labor. The problem I brought up is still there, just in the hands of the publisher, and not the writer. Now the publisher gets to profit from work they did not do, and you are lucky to get a good cut.
Your labor has value to everyone else, even if that value is speculative. If we don't have a mechanism to commoditize the ends of labor, then we can just speculate instead. Speculating the value of labor is more uncertain than valuing copyrighted works, which means that the business of labor speculation cannot compete with the business of copyright valuation. At the same time, copyright is a lie: the "product of a work" is a totally arbitrary imagined boundary that can't always be meaningfully drawn in the first place; meaning that entire categories of work are impossible to copyright at all. Removing that lie would put everyone on a level playing field, where all labor is valuable, and all valuation is fair.
> We should get rid of copyright and patents... Value labor directly.
Ireland's UBI for artists seems like the only real solution that gets to the heart of the problem
In effect the source code is being copied by the LLM. This is what it's designed to do. LLMs are a lossy statistical compression of their training data.
If you give it a prompt telling it to replicate a product that's in its training set then its optimal next token prediction output is going to be to a lossy copy of that product's source code.
There are no novel UIs, so copying UIs is okay, and necessary. As for source code, I'm a stickler for the license. The modern set of licenses cover any scenario I can think of, relatively fairly. AI is merely a tool, so the craftsman still owns the output. If the output violates a license, then the craftsman should be held to account.
I’ve seen companies get sued and lose in court for copying UIs. So no, legally speaking, copying a UI is not something that companies ignore.
The look and feel of an interface could be covered under a design patent, which would be a different question than copyright, and a much higher bar to enforce (particularly, protection is far from automatic).
You’re talking about enforcement which is a fair point of view, but I’m taking about likelihood of getting sued and having to settle in order to avoid the risk of losing a case in court. Which is the more common case for situations where one company alleges the other stole their IP.
Can you point to cases so I can google them?
Borland Quattro.
In that case no infringement was found though?
I'd argue that software is an "applied art", and needs a "high threshold of originality" to be protected.
https://en.wikipedia.org/wiki/Threshold_of_originality
Oh and if it's not human generated, you can just copy it.
I think there are couple things going on here
The replication/copying has always been there in one form or another. The bar has traditionally been higher for reputation and monetary risks.
Lately the legal bar is the one that going down, ease of replication makes it even more tempting and when big players are doing it at scale (bots) then it validates the strategy in one way or another.
If anything, there have to be downstream consequences of this with time, libraries to pollute the front end code for LLMs are most likely going to get popular and probably one way to make it harder for your IP to be replicated.
Think it'll be hard to define in any sufficiently specific manner legally because UI/UX/source/functionality aren't entirely separate, but ethically I reckon it comes down to what one means by copy UI.
e.g. If you're creating an uptime dashboard...they all kinda look the same anyway and there aren't that many ways to do it so that seems OK. If it's copying an comprehensive UI with layout and flow between the various pages etc then you're getting a bit closer to theft.
That is why we have judges, to make judgement calls. Its probably going to come down to how closely you are copying.
It depends. Is the text auto-generated from a framework? Is it creative? Is it worthy of copyright? It seems more instructional.
I say let them sue for copyright infringement for the text. Let them sue for breaking a license. Let's see how it works out. Doesn't impact me and if it costs the rich money, good. Let them suffer.
copyright is only as good as the enforcement. enforcement is exhausting and expensive.
Enforcement is well within the capabilities of major orgs (ie nintendo) who litigate internationally
So to be clear the answer is emphatically "no". If you copy everything else, the defense that the source code is technically different will not save you.
copying other business pixel-to-pixel and direct claims agains competitor is just distasteful. add to this recent YC "AI that MITM API and re-implments anything automatically" is very bad image to YC.
with such bad behavior from SWE community, you just got to lock down your app behind certificate pinning, hardware attestation, gRPC/protobufs, and internal data only. no more "free open web in browsers" when you get gents like this stealing other peoples efforts.
By that logic OpenOffice would infringe on Microslop Office because "it looks the same" (as an older version of M$ Office).
A sad state of affairs when the law is what people look to in order to decide between right and wrong.
No definitely not. I've never seen a patent include code. They're more likely to describe IP in a work flow diagram.
How do patents relate to this copyright question?
I made a leap there without explaining. You can copyright code the same way you copyright a book. But in doing so you protect the text but not inherently the IP. Copyrighted code has a whole "originality" argument in it's defensibility. Patents do too, but they would never patent the code itself. If you want to protect yourself... it's belt and suspenders.
There was an article here about stealing the other day and only change 3%, so I guess it’s working already!
The clean room PC compatible BIOS's were written that way for a reason.
we need open training data models, and not just open weights.
if you cant prove that the source code wasnt trained on, how can you show that its not a copy of the copywrited original?
not a copyright issue
there may some other intellectual property remedy, or not, but it isn't copyright
hope that helps
technically may be it is. practically you are a dick.
(e.g. instagram copying snapchat)
...hopefully as other comments said, that LLMs make us abandon the confusing idea of "intellectual property" so we don't have to ask questions like this or get in to torturous questions if this or that thing is "infringing" or not
There's a couple of related issues being conflated here, and I'm not sure which one to bring up, mainly because I'm not sure in what direction the copying would be ruled to have gone. So I'll just mention all the cases.
The first thing to note is that nonliteral copying can still be infringing. Actually, among copyright cases that actually go to trial, most of them are not bit-exact matches ("striking similarity" in legalese). The lower standard those cases would have to meet is substantial similarity, which requires proving both access and similarity. In other words, in order for you to produce a copy[0], you have to have both seen the original and produced something that is close enough if you squint at it.
So let's say Papermark is the original and Corgi Dataroom copied it. If this actually went to trial, a significant amount of discovery would be spent harvesting all the e-mails and messages Corgi's development team sent to one another. Any evidence of knowledge or access to Papermark would probably be enough to prove a copyright violation.
You mentioned LLMs, and some of the tweets here also mention them. I have no clue if either product used an LLM, but it's important to note that in the US, anything written by an LLM does not accrue copyright protection. So, if Papermark was LLM-authored, as a threshold matter, they would have to register[2] a very specific copyright that neatly carves out the LLM-authored bits. The judge would then only consider the parts of the code with verifiable human authorship, which would severely weaken Papermark's case.
In the reverse case - i.e. Corgi Dataroom is LLM-authored - then Papermark's case becomes stronger. Note how I didn't say "AI slop is public domain" last paragraph, because it isn't. There are still unsettled legal questions as to whether or not training on copyrighted works is legal and if using an LLM trained on that work constitutes access to it. Furthermore, LLMs can have search tools that would give them access to code not within the training set, which would also be a more straightforward copyright violation. So if it turns out Dataroom's developers are all Claude fans, and Claude is copying Papermark code, then it's just a normal license violation. LLMs do not launder copyright.
We can also consider the case where BOTH tools are LLM-authored. In this case, there might just not be a copyright case at all. Technically speaking, there would be some third class of plaintiffs who have been infringed, but they would have to choose to sue. It is a long-standing principle in law that you are not allowed to sue for other people's harms[1]. So in this case, nobody would have a case.
> Now software developers are feeling what authors and artist felt
It was specifically the FOSS community that sounded the alarm about training data theft first, because the FOSS community correctly understood coding agents to be an attack on copylefts[3] (albeit through the incorrect belief that LLMs could launder away copyright interest, as opposed to it just being told to make a noninfringing substitute).
[0] I am skipping over notions of fair use and derivative works as they would complicate the analysis and do not apply here.
[1] In general, the operating principle of American courts is "fuck around and find out", and this implies that the court is only allowed to find out once someone has fucked around. Otherwise, the courts could just sue themselves to rule on whatever the hell they want. Isn't adversarial common law GREAT!?
[2] Yes, copyright registration is mandatory in the US, otherwise you can't sue, which is the whole point of copyright. The Berne Convention only half-applies here.
[3] Clauses in licenses that require modifications to the work to be provided under the same license terms. Creative Commons calls these Share-Alike licenses.
> So if it turns out Dataroom's developers are all Claude fans, and Claude is copying Papermark code, then it's just a normal license violation. LLMs do not launder copyright.
That's an easy case, but a more interesting question to me is if there wasn't any direct source code access (similar to an open source example I can't remember the name of atm). If you give an LLM requirements matching the copyrighted products capabilities, and it doesn't have Internet access, and it spits out different code that accomplishes the same stuff papermarks product does, do you believe courts would allow that, and copyrighted could be laundered legally?
If, in fact, you copy computer software by spec and never had access to the code -- that's a clean room reimplementation and doesn't violate copyright. Copyright covers the human expression of software -- the text of the code itself. It can also cover non-code assets included in your software like graphics. Copyright does not cover the ideas, goals, designs, concepts, principles, etc of your software.
Did you agree to terms of use? Did you have to click a check box that you agree to terms of use before seeing or having access to the items you copied? Click wrap. If in the contract that you agreed to there is language that you agreed to not copy the work, then you likely are in breach of contract. If it is publicly available knowledge probably not breach of contract. I’m not a lawyer of course.
Copyright applies even without the contract.
There is a lot of copying that isn't protected by copyright. It is possible to include what can't be copied that might not fall under copyright in the terms of service contract. Many people not being able to use copyright to prevent copying instead successfully sued based on breach of contract.
Companies should understand that they can protect their IP this way.