The conversation around ownership and AI is less about absolutes and more about clarity and accountability.
Ooh, free trash code!
Stick to the GPL licensencing of your code whenever possible and the garbage EEE can’t subdue you. (Embrace extend exthinguish.)
If they plagiarize it they kinda ow you the honor.
Hower, plagiarism is still plagiarism, so you better actually write some of your code by hand.
The the greatest intellectual property theft and laundering scheme in human history
That sounds like complete bullshit to me. Even if the logic is sound, which I seriously doubt, if you use someone’s code and you claim their license isn’t valid because some part of the codebase is AI generated, I’m pretty sure you’ll have to prove that. Good luck.
There was a case in which a monkey took a picture and the owner of the camera wanted to publish the photo. Peta sued and lost because an animal can’t hold any copyright as an human author is required for copyright.
https://en.wikipedia.org/wiki/Monkey_selfie_copyright_dispute
As you also find in the wikipedia article, this case is used to argue that ai generated content is not by an human author and consequently not copyrightable.
I’d argue that this is a different scenario, as AI is a tool, not a being. At least at this point.
A complex tool, but really just a tool. Without the human input, it can’t do shit.
Exactly. If I use online Photoshop or whatever, and I use the red eye removal tool, I have copyright on that picture. Same if I create a picture from scratch. Just because someone like OpenAI hosts a more complex generator doesn’t mean a whole new class of rules applies.
Whomever uses a tool, regardless of the complexity, is both responsible and benificiary of the result.
I work for a large enterprise firm, our corporate lawyer has told be about this exact scenario so I’m inclined to believe it’s real.
That being said, for established projects it won’t be that hard to prove the non-AI bit because you have a long commit history that predates the tooling.
Even if you were to assume that all commits after a certain date were AI generated, the OP is slightly off in their attestation that any AI code suddenly makes the whole thing public domain, it would only be if a majority of the codebase was AI coded (and provably so).
So yes all the vibe coded shite is a lost cause, but stuff like Windows isn’t in any danger.
I’ll stay sceptical until there is court cases supporting this logic as a precedent.
I think that’s actually quite sensible, our lawyer wasn’t flagging some clear cut legal certainty, he was flagging risk.
Risk can be mitigated, even if the chance of it panning out is slim.
If there was an actual civil suit you’d probably be able to subpoena people for that information, and the standard is only more likely than not. I have no idea if the general idea is bullshit, though.
IANAL
You forgot the heart
I ♥️ ANAL
Would that be North African Lawyer, or North American Lawyer?
In any case, we’re splitting the cheque. /s
Are you suggesting that lawyers migrate?
This whole post has a strong ‘Sovereign Citizen’ vibe.
I do not give Facebook or any entities associated with Facebook permission to use my pictures, information, messages, or posts, both past and future.
The Windows FOSS part, sure, but unenforceable copyright seems quite possible, but probably not court-tested. I mean, AI basically ignored copyright to train in the first place, and there is precedent for animals not getting copyright for taking pictures.
If it’s not court tested, I’m guessing we can assume a legal theory that breaks all software licensing will not hold up.
Like, maybe the code snippets that are AI-made themselves can be stolen, but not different parts of the project.
That seems a more likely outcome.
As it should. All the idiots calling themselves programmers, because they tell crappy chatbot what to write, based on stolen knowledge. What warms my heart a little is the fact that I poisoned everything I ever wrote on StackOverflow just enough to screw with AI slopbots. I hope I contributed my grain of sand into making this shit little worse.
Do it in a way that a human can understand but AI fails. I remember my days and you guys are my mvp helping me figure shit out.
Most “humans” don’t understand reality. So you’re postulative challenge invention isn’t going find a break you seek to divine. Few exist. I’m yet to find many that can even recognize the notion that this language isn’t made to mean what think you’re attempting to finagle it into.
Evil Money Right Wrong Need…
Yeah…I could go on and on but there’s five sticks humans do not cognate the public consent about the meaning of Will Never be real. Closest you find any such is imagination and the only purpose there is to help the delirious learn to cognate the difference and see reality for what it may be.
Good fucking luck. Half the meat zappers here think I am an AI because break the notion of consent to any notion of a cohesive language. I won’t iterate that further because I’ve already spelt out why.
That’s not what that research document says. Pretty early on it talks about rote mechanical processes with no human input. By the logic they employ there’s no difference between LLM code and a photographer using Photoshop.
By that same logic LLMs themselves (by now some AI bro had to vibe code something there) & their trained datapoints (which were on stolen data anyway) should be public domain.
What revolutionary force can legislate and enforce this?? Pls!?
By that same logic LLMs themselves (by now some AI bro had to vibe code something there)
I’m guessing LLMs are still really really bad at that kind of programming. The packaging of the LLM, sure.
& their trained datapoints
For legal purposes, it seems like the weights would be generated by the human-made training algorithm. I have no idea if that’s copyrightable under US law. The standard approach seems to be to keep them a trade secret and pretend there’s no espionage, though.
The packaging of the LLM, sure.
Yes, totally, but OP says a small bit affects “possibly the whole project” so I wanted to point out that includes prob AIs, Windows, etc too.
Aren’t you all forgetting the core meaning of open source? The source code is not openly accessible, thus it can’t be FOSS or even OSS
This just means microslop can’t enforce their licenses, making it legal to pirate that shit
It’s just the code that’s not under copyright, so if someone leaked it you could legally copy and distribute any parts which are AI generated but it wouldn’t invalidate copyright on the official binaries.
If all the code were AI generated (or enough of it to be able to fill in the blanks), you might be able to make a case that it’s legal to build and distribute binaries, but why would you bother distributing that slop?
Even if it were leaked, it would still likely be very difficult to prove that any one component was machine generated from a system trained on publicly accessible code.
That’s not even remotely true…
The law is very clear that non-human generated content cannot hold copyright.
That monkey that took a picture of itself is a famous example.
But yes, the OP is missing some context. If a human was involved, say in editing the code, then that edited code can be subject to copyright. The unedited code likely cannot.
Human written code cannot be stripped of copyright protection regardless of how much AI garbage you shove in.
Still, all of this is meaningless until a few court cases happen.
So wait, if my start up does a human written “Hello World” and the rest is piled on AI slop it can’t be stripped of copyright? Or is “Hello World” too generic to hold a copyright at all?
Granted, as you said this all has to be defined and tested in court, I’m just trying to understand where the line as you see it is.
https://www.reinhartlaw.com/news-insights/only-humans-can-be-authors-of-copyrightable-works
https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf
A human must be involved in the creation. A human can combine non-human created things to make something new, but the human must be involved, and the non-human created elements likely lack protection themselves.
“Hello World” is prior art.
Just add it after ai writes everything else then.
deleted by creator
People will believe anything if the icon on the tweet looks authoritative and the grammar is sound.
As much as I wish this was true, I don’t really think it is.
It’s just unsettled law, and the link is basically an opinion piece. But guess who wins major legal battles like this - yep, the big corps. There’s only one way this is going to go for AI generated code
Worst case is that its the owner of the agent that recieves the copyright, so all vibe coded stuff outside local ai will be claimed by the big corpos
I actually think that’s the best case because it would kill enterprise adoption of AI overnight. All the corps with in-house AI keep using and pushing it, but every small to medium business that isn’t running AI locally will throw it out like yesterday’s trash. OpenAI’s stock price will soar and then plummet.
The big AI companies would just come out with a business subscription that explicitly gives you copyright.
Damn
Unlikely since, as you say, it would deter business. OpenAI already assigns rights of output to the end user according to their licensing and terms.
No attempt to argue with you, personally is intended here. But your comment raises another question that I’m not sure the law has answered yet.
What rights does OpenAI have in the output of ChatGPT in the first place? Because if the answer is “Not much” then their transfer of rights to the output to the user doesn’t necessarily mean much.
After all, OpenAI can only transfer rights that they have. If they don’t have any to begin with… 🤷♂️
Yep, totally fair question, and one that’s being tested legally on many fronts. Rulings are generally siding with AI companies on the training side (using copyrighted works to train models is fair use) but there aren’t many decisions yet about output. The next few years will be interesting.
It is true that AI work (and anything derived from it that isn’t significantly transformative) is public domain. That said, the copyright of code that is a mix of AI and human is much more legally grey.
In other work, where it can be more separated, individual elements may have different copyright. For example, a comic was made using AI generated images. It was ruled that all the images were thus public domain. Despite that, the text and the layout of the comic was human-made and so the copyright to that was owned by the author. Code, obviously can’t be so easily divided up, and it will be much harder to define what is transformative or not. As such, its a legal grey area that will probably depend on a case-by-case basis.
Yeah, it’s like products that include FOSS in them, only have to release the FOSS stuff, not their proprietary. (Was kind of cute to find the whole GNU license buried in the menus of my old TiVo…)
So, you’re telling me I can copypaste 100% some of the ai slop books on amazon and resell it as mine? Brb, gonna make a shit site an absolute diarrhea
If the AI generated code is recognisably close to the code the AI has been trained with, the copyright belongs to the creator of that code.
Since AIs are trained on copyleft code, either every output of AI code generators is copyleft, or none of it is legal to use at all.
I may be wrong but I think current legal understanding doesn’t support this
Under U.S. law, to prove that an AI output infringes a copyright, a plaintiff must show the copyrighted work was “actually copied”, meaning that the AI generates output which is “substantially similar” to their work, and that the AI had access to their work.[4]
I’ve found a similar formulation in a official German document before posting my above comment. Essentially, it doesn’t matter if you’ve
“stolen”copied somebody else’s code yourself and used it in your work or did so by using an AI.
Even if it were, it would be for you or I, but not for Microsoft, apple, Google, or Amazon.
The US copyright office confirms this. https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf
The use of AI tools to assist rather than stand in for human creativity does not affect the availability of copyright protection for the output. Copyright protects the original expression in a work created by a human author, even if the work also includes AI-generated material.
I’m not sure where you get that from, I’m pretty sure vibe coding still complies with these indications
“AI-generated” works can be copyrighted. However, on the condition that the AI-generated elements are explicitly mentioned in the “Excluded Material” field. In other words, the parts generated by AI are not protected, only the parts that are expressed by human creativity. Courts in the U.S have already rejected registration for many AI works because of that. Regardless, it’s still a contentious matter.
P.S. I am completely opposed to (generative) AI as well as the copyright system. I’m just stating my findings researching the law and court cases.
As mentioned elsewhere in this thread it won’t matter either way unless tested in court and that will never happen for most companies.
Did you even read your own report? It says that AI works are copyrightable in certain circumstances, not that they make a whole project public:
Copyright law has long adapted to new technology and can enable case-by- case determinations as to whether AI-generated outputs reflect sufficient human contribution to warrant copyright protection. As described above, in many circumstances these outputs will be copyrightable in whole or in part—where AI is used as a tool, and where a human has been able to determine the expressive elements they contain. Prompts alone, however, at this stage are unlikely to satisfy those requirements.
“AI-generated” works can be copyrighted. However, on the condition that the AI-generated elements are explicitly mentioned in the “Excluded Material” field. In other words, the parts generated by AI are not protected, only the parts that are expressed by human creativity. Courts in the U.S have already rejected registration for many AI works because of that.
P.S. I am completely opposed to (generative) AI as well as the copyright system. I’m just stating my findings researching the law and court cases.
The part that is untrue is the “public domain” part. If you generate code then you don’t own it because the actual human work that went into creating it was done by the owner of the AI Model and whatever they trained on.
Iirc it’s even funnier: the relevant case law comes from Naruto v Slater. A case about a monkey taking a selfie and a photographer failing to acquire copyright of it (https://en.wikipedia.org/wiki/Monkey_selfie_copyright_dispute).
The copyright belonged to whoever shot the selfie, but because it was the monkey and animals aren’t juristic entities, they can not hold copyright. Therefore, as it stands and as new case law outlines, AIs are compared to monkeys, in that the copyright would fall onto them but it’s not a juristic entities either, and therefore copyright just vanishes and no one can claim it.
The wikipedia page suggests current cases on generative AI directly build on this.
It was an especially interesting case because there was a question of whether the photographer lied about who actually took the picture. So he could either claim the monkey took it an lose the copyright or claim he took it and have it lose all value.
See that’s kind of what I’m talking about. The monkeys who pressed the buttons to make the AI Generate the code isn’t the computer, it isn’t the user, it’s the employees at the AI company. My advice is that until laws are properly in place that we shouldn’t use AI for any generative industry.
The AI company didn’t do shit. They stole apples from someone elses tree and threw it in a blender. They didn’t make the apples, nor did they buy them, so they don’t legally own the juice.
While I agree with the general idea, please don’t call piracy “stealing”. It’s not stealing, whether you do it or some giant corpo.
While I agree², their use of “steal” makes sense in the analogy because the apple doesn’t belong to the “thief”; besides, you can’t pirate an apple
Evidently you can!
“Apple piracy is a problem,” said Lynnell Brandt, chief executive officer of Proprietary Variety Management, the Yakima firm hired to handle the Cosmic Crisp rollout. “Some are propagating trees without permission. Some are stolen in the middle of the night.”
Don’t call copyright infringement “piracy,” either.
If the AI produces verbatim the licensed works of others then the others own it.
If the AI took individual unlicensed elements and pieced them together then the AI Company owns it.
In any and every case, neither the User nor the Public Domain owns it. Moral of the story is: never use AI for anything.
The AI company stole other people’s code, threw it into a blender, and is selling the output. They didn’t do any real work, and they don’t own the materials. They have no legal claim over the result. You do not own a car you made from stolen parts, no matter how many cars you stole from.
Stop trying to imply your buddies at AI companies have value.
We appear to be talking in circles.
I’m literally sitting here telling people it isn’t safe to use AI Code, you’re doing the opposite, and you’re accusing me of being buddies with the Slop Companies?
You’re just making shit up. The US Court of Appeals for the DC Circuit has affirmed that AI-generated work is in the public domain. Put up or shut up.
Edit: Additionally, the US Copyright Office writes:
As the agency overseeing the copyright registration system, the [Copyright] Office has extensive experience in evaluating works submitted for registration that contain human authorship combined with uncopyrightable material, including material generated by or with the assistance of technology.
How does this work in practice? Someone would have to prove that it’s AI generated, which isn’t straight forward.
Also, I’m not clear this protects the release of code centered a trade secret or under NDA.
So while the court ruled it’s public domain. Could it still be prevented from release? Like a Microsoft employee couldn’t just dump sections of the AI code to the internet I imagine.
https://www.upcounsel.com/patents-trademarks-copyrights-and-trade-secrets
Competitive advantage: Trade secrets can cover information that would not qualify for patents or copyright but still has economic value.
I would imagine dumping Microsoft code to the internet would be sued under NDA
The answer is that it’s messy and that I’m not qualified to say where the line is (nor, I think, is anyone yet). The generated parts are not copyrightable, but you can still have a valid copyright by bringing together things that aren’t individually copyrightable. For example, if I make a manga where Snow White fights Steamboat Willie, I’ve taken two public domain elements and used them to create a copyrightable work.
So it’s not like the usage of AI inherently makes a project uncopyrightable unless the entire thing or most of it was just spat out of a machine. Where’s the line on this? Nobody (definitely not me, but probably nobody) really knows.
As for courts ever finding out, how this affects trade secret policy… Dunno? I’m sure a Microsoft employee couldn’t release it publicly, because as you said, it’d probably violate an NDA. If there were some civil case, the source may come out during discovery and could maybe be analysed programmatically or by an expert. You would probably subpoena the employee(s) who wrote the software and ask them to testify. This is just spitballing, though, over something that’s probably inconsequential, because the end product is prooooobably still copyrightable.
This kind of reminds me of the blurry line we have in FOSS, where everyone retains the copyright to their individual work. But if push comes to shove, how much does there need to be for it to be copyrightable? Where does it stop being a boilerplate
forloop and start being creative expression?
Is letting an LSP auto complete for you “assistance of technology”?
It begins by asking “whether the `work’ is basically one of human authorship, with the computer [or other device] merely being an assisting instrument, or whether the traditional elements of authorship in the work (literary, artistic, or musical expression or elements of selection, arrangement, etc.) were actually conceived and executed not by man but by a machine.” [23]
In the case of works containing AI-generated material, the Office will consider whether the AI contributions are the result of “mechanical reproduction” or instead of an author’s “own original mental conception, to which [the author] gave visible form.” [24]
The answer will depend on the circumstances, particularly how the AI tool operates and how it was used to create the final work.[25] This is necessarily a case-by-case inquiry.
That’s the rest of what you posted. I guess you just didn’t read it, right? Even if it comes right after and is part of the same paragraph. What a joke.
I clarified this a bit in a follow-up comment, but my first comment was simplifying for the sake of countering:
[it’s not in the public domain] because the actual human work that went into creating it was done by the owner of the AI Model and whatever they trained on.
Their claim that the copyright for AI-generated works belongs to the model creator and the authors of the training material – and is never in the public domain – is patent, easily disprovable nonsense.
Yes, I understand it’s more nuanced than what I said. No, it’s not nuanced in their favor. No, I’m not diving into that with a pathological liar (see their other comments) when it’s immaterial to my rebuttal of their bullshit claim. I guess you just didn’t read the claim I was addressing?
Technology is an extremely vague word in this context. If the US Court of Appeals for the DC Circuit has affirmed that then I haven’t heard of it, it’s not posted here, and most importantly: such rules are not currently enshrined in law.
Removed by mod
Removed by mod
No it’s that you’re trying to walk back a provably false claim and then deflect the claims by pretending the people calling you out are doing so because they like AI instead of, you know, valuing the truth.
I walk back no claims. The AI Companies have more claim on ownership of the output than the public. Don’t use Slop Code, it’s not safe.
Counterpoint: how do you even prove that any part of the code was AI generated.
Also, i made a script years ago that algorithmically generates python code from user input. Is it now considered AI-generated too?
i made a script years ago that algorithmically generates python code from user input. Is it now considered AI-generated too?
No, because you created the generation algorithm. Any code it generates is yours.
Not how I understand it, but I’m not a lawyer. The user that uses the script to generate the code can copyright the output and oop can copyright their script (and the output they themself generate). If it worked like you said, it would be trivial to write a script that generates all possible code by enumerating possible programs, then because the script will eventually generate your code, it’s already copyrighted. This appear absurd to me.
If the script copies chunks of code under the copyright of the original script writer, I typically see for those parts that the original owner keeps copyright of those chunks and usually license it in some way to the user. But the code from the user input part is still copyrightable by the user. And that’s that last part that is most interesting for the copyright of AI works. I’m curious how the law will settle on that.
I’m open to counterarguments.
While nobody created neural nets and back propagation
Computer output cannot be copyrighted, don’t focus on it being “AI”. It’s not quite so simple, there’s some nuance about how much human input is required. We’ll likely see something about that at some point in court. The frustrating thing is that a lot of this boils down to just speculation until it goes to court.
OP is obviously ignorant of how much tooling has already helped write boiler plate code.
Besides AI code is actually one of the things that’s harder to detect, compared to prose.
And all that said, AI is doing an amazing job writing a lot of the boilerplate TDD tests etc. To pretend otherwise is to ignore facts.
AI can actually write great code, but it needs an incredibly amount of tests wrapped around and a strict architecture that it’s forced to stick to. Yes, it’s far too happy sprinkling magic constants and repeat code, so it needs a considerable amount of support to clean that up … but it’s still vastly faster to write good code with an AI held on a short leash than it is to write good code by hand.
Guess you can’t really prove that, unless you leave comments like “generated by Claude” in it with timestamp and whatnot 😁 Or one can prove that you are unable to get to that result yourself.
So nonsense, yes.
Or one can prove that you are unable to get to that result yourself.
Oh shit… I’ve got terabytes of code I’ve written over the years that I’d be hard-pressed to even begin to understand today. The other day I discovered a folder full of old C++ libraries I wrote 20+ years ago, and I honestly don’t remember ever coding in C++.
There is absolutely no way you wrote terabytes of code lmao.
True enough, and I expected to get checked on that.
Regardless… along with the archives, assets and versioned duplicates, my old projects dating back to the 90s somehow now fill multiple TB of old hard drives that I continue to pack-rat away in my office. Useless and pointless to keep, but every piece was once a priority for someone.
Cursor, an ai/agentic-first ide, is doing this with a blame-style method. Each line as it’s modified, added DOES show history of ai versus each human contributor.
So, not nonsense in probability, but in practice – no real enforcement to turn the feature on.
Why would you ever want this?
If you pushed the bug that took down production - they aren’t gonna whataboutism the AI generated it. They’re still going to fire you.
It makes little difference IMHO. If you crash the car, you can’t escape liability blaming self driving.
Likewise, if you commit it, you own it, however it’s generated.
It’s mainly for developers to follow decisions made over many iterations of files in a code base. A CTO might crawl the gitblame…but it’s usually us crunchy devs in the trenches getting by.
Sorry, but as another reply: pushing bugs to production doesn’t immediately equate to firing. Bug tickets are common and likely addressing issues in production.
Hence the “took down production”
I guess you mean like full outtage for all users? My bad just a lot of ways to take the verb “down” for me. Still, though, what a crappy company to not learn but fire from that experience!
Uh, yes, that’s what they call a generative ai
I think, to punish Micro$lop for its collaboration with fascists and its monopolistic behavior, the whole Windows codebase should be made public domain.
does the public really want more garbage than they already has?
Do you not want all the hardware support Linux is missing to suddenly become available?
The kernel and NTFS seem decent from what i heard. Or at least was (the kernel, no guess what they vibecoded into it now).
About NTFS: it was actually pretty good for it’s time (90s), but the tooling makes no use of some of it’s better features and abuses some others close to breaking point. Literally pearls for the sows.
There were decent (at least, worked for me) NTFS drivers for Linux like 20 years ago. (Back when I felt the need to dual boot)
We don’t have to use it for anything other than compatibility.
Personally I’d very much like for ACPI to be un-fucked
Haven’t we been exposed to enough horror?
Probably wouldn’t work, at this point I doubt they’re capable of feeling shame anymore, if they ever were.
I had a similar thought. If LLMs and image models do not violate copyright, they could be used to copyright-wash everything.
Just train a model on source code of the company you work for or the copyright protected material you have access to, release that model publicly and then let a friend use it to reproduce the secret, copyright protected work.
btw this is happening actuallt AI trained on copyrighted material and it’s repeating similar or sometimes verbatim copies but license-free :D
This is giving me illegal number vibes. Like, if an arbitrary calculation returns an illegal number that you store, are you holding illegal information?
(The parallel to this case is that if a statistical word prediction machine generates copyrighted text, does that make distribution of that text copyright violation?)
I don’t know the answer to either question, btw, but I thought it was interesting.
In case of illegal numbers, intention matters. Because any number could be converted to different numbers, for instance through ‘xor encryption’ different ‘encoding’ or other mathematical operations, which would equally be illegal if used with the intention to copy copyright protected material.
This was the case previously. You cannot simply reencode a video, a big number on your disk, with a different codec into another number in order to circumvent copyright.
However, if big business now argues that copyright protected work encoded in neuronal network models is not violating copyright and generated work has no protection, then this previous rule isn’t true anymore. And we can strip copyright from everything using that ‘hack’.
