AI written books

[DiapDealer]

I'm not seeing anything in that easy-to-understand analysis that definitively decrees that DRM removal = Copyright infringement. There have been court cases that have come down on either side of that particular argument. Meaning that while the two can certainly be intertwined, DMCA violation (which DRM removal falls under) is not legally synonymous with copyright violation. One may lead to the other, but they are separate.

[j.p.s]

Just noting that the C in DMCA is copyright. (Not claiming that anything in the DMCA makes sense.)

[DiapDealer]

Originally Posted by j.p.s

Just noting that the C in DMCA is copyright. (Not claiming that anything in the DMCA makes sense.)

Yep. But violating the DMCA's circumvention law (while still being illegal in many jurisdictions) has been considered separate from copyright infringement/violation by some courts.

I am merely taking issue with @nabsltf's claim that merely defeating the DRM is a "copyright violation all by itself." I don't believe the article they linked to supports that claim.

[Quoth]

No, removing or defeating DRM on its own isn't copyright violation. It's what you do after that which either violates or doesn't violate copyright.

The USA DMCA is a completely separate thing, despite having "copyright" in the name.

[DiapDealer]

That was my take too. Thanks.

[Quoth]

If you take copyright content from pirate sites and "some how reshare it" is it copyright violation?
DRM arguments are irrelevant?. Source is irrelevant? It's what you do with the content that will count in court?

The suits claim the authors’ works were obtained from “shadow library” sites that have “long been of interest to the AI-training community”.

https://www.theguardian.com/technolo...t-infringement

[DiapDealer]

Originally Posted by Quoth

If you take copyright content from pirate sites and "some how reshare it" is it copyright violation?
DRM arguments are irrelevant?. Source is irrelevant? It's what you do with the content that will count in court?

"Some how reshare it" is going to need to be defined. It's either being reshared, or it isn't. If "somehow" means anything other than sharing the downloaded original in its entirety, then the answer to whether or not it's copyright violation is maybe/maybe not.

If you mean "shared" as in new fiction created by systems trained on the originals, then my guess is probably not.

But otherwise, yes. I think it's the latter: "It's what you do with the content that will count in court."

In the case of chatGPT and others using pirated downloads/scrapes to train their algorithms, I find the question of copyright to be moot. If it can be proved they used (or are using) pirated downloads/scrapes to train their algorithms, then they should be prosecuted for mass piracy.

If they (or anyone) used legally obtained (DRM irrelevant) copies to train their systems, I don't see a copyright problem either. Those copies are not being disseminated (with or without permission).

[Hitch]

Originally Posted by DiapDealer

"Some how reshare it" is going to need to be defined. It's either being reshared, or it isn't. If "somehow" means anything other than sharing the downloaded original in its entirety, then the answer to whether or not it's copyright violation is maybe/maybe not.

If you mean "shared" as in new fiction created by systems trained on the originals, then my guess is probably not.

But otherwise, yes. I think it's the latter: "It's what you do with the content that will count in court."

In law, unless you're dealing with a very unusual court, intent always, always matters. It may not be the decider, mind you, between guilty/not-guilty, but it can certainly affect the damages awarded, etc.

If you make a copy for yourself, that's one thing. I can't imagine any loony being crazy enough to chase/sue you for that. So...whatever.

If you make a copy for 100K people, that's different. That clearly signals "intent" and that's a whole other kettle.


Too many variables and "what ifs" here. We need to settle on some limits so that we can start to discuss it intelligently and yes, I'm painfully well aware that this is precisely the problem in discussing it--too many unknown unknowns and all that.

Hitch

[nabsltd]

17 USC is all about copyright. So, if somebody raises a lawsuit under chapter 12, it's a copyright lawsuit.

Yes, courts are all over the place about DRM removal, but it's pretty clear that since eBooks are allowed to be copyrighted¹, that means that removing DRM by itself creates a copy not authorized by the creator, thus it's also a violation of 17 US § 106(1).

Again, I'm not here making moral or ethical judgments...I'm just explaining what the US law says, as written. And, I know about various court rulings, but the #1 thing is that there has never been (and likely never will be) is a lawsuit about nothing but DRM removal. If that's all that happens, the copyright holders won't sue because they don't want to lose and set the precedent that "it's OK for personal use". You don't even see the manufacturers of DRM solutions raising the suits, either, even though you lawsuits of other reverse-engineering filed under 17 USC 1201.

¹According to the exact letter of the law, they might not be, since they aren't fixed in a material nor are they computer programs. Everything that is not a computer program that is streamed or downloaded falls into this hole of non-definition in the written law. It's only court decisions that have codified that these items have copyright protection, and there still hasn't been an umbrella USSC decision that unifies all districts in this. There has been a bit of an update to the definitions in section 101, but words that describe "storage on a computer" still aren't there.

[DiapDealer]

Originally Posted by nabsltd

Yes, courts are all over the place about DRM removal, but it's pretty clear that since eBooks are allowed to be copyrighted¹, that means that removing DRM by itself creates a copy not authorized by the creator, thus it's also a violation of 17 US § 106(1).

No. It's not clear. No matter how much you want it to be clear, it's not. That's why courts are all over the place about DRM removal. Hence why no one can definitively state that DRM Removal == Copyright Infringement/Violation. Which is what you tried to definitively state was true in your first post on the subject.

You don't have to keep up the moral judgement disclaimers, by the way. I don't think anyone here is coming at this from a moral standpoint.

[Quoth]

Originally Posted by j.p.s

Just noting that the C in DMCA is copyright. (Not claiming that anything in the DMCA makes sense.)

And D in these is Democratic:
GDR (now gone)
DPRK (Democratic People's Republic of Korea, North Korea)
DUP (Democratic Unionist Party: two out of 3 ain't bad.)

The Meta's LLama and MS/Open AI's Chat usage will argue "Fair Use". However that's never been formalised in most countries other than USA. It will be interesting to see the arguments. I'd expected Google to lose the Scanning case they won, but the Court believed Google/Alphabet's claims.

[Quoth]

Originally Posted by nabsltd

Yes, courts are all over the place about DRM removal, but it's pretty clear that since eBooks are allowed to be copyrighted¹, that means that removing DRM by itself creates a copy not authorized by the creator, thus it's also a violation of 17 US § 106(1).

No, it's not. Otherwise you couldn't listen / view / read the content.

[nabsltd]

Originally Posted by Quoth

No, it's not. Otherwise you couldn't listen / view / read the content.

When you read or view content protected by DRM on a device approved by the copyright holder, you are not "removing" or "defeating" DRM, so that does not apply.

[Quoth]

The copyright holder isn't explicitly approving Kindles, Kobos or Nooks.

[DNSB]

Originally Posted by nabsltd

When you read or view content protected by DRM on a device approved by the copyright holder, you are not "removing" or "defeating" DRM, so that does not apply.

Actually, you are removing the DRM otherwise what you see on the device screen would be undecipherable garbage.