[chrismorgan]

> It’s the safest option, because at a certain scale, it is impractical for large platforms to evaluate the validity of every single takedown request they receive.

It has long seemed crazy to me that, as a society, we’ve allowed large companies to argue that they can’t do basic things that their smaller competitors have to. Provide customer support. Assess legal challenges. Et cetera. Should we not rather say: you have the resources, use them! This is a cost of getting big. You have economies of scale in other areas, don’t try to evade responsibility here.

[SR2Z]

> Assess legal challenges.

There will be nothing but pain and frustration if you ask corporations to try and supplant the courts. Copyright law is old and does not make provisions for the modern era.

You get a copyright when you create a work and it does not require any kind of registration. Establishing who has a copyright, if the work is copyrightable in the first place, or if an alleged infringement is fair use or not are thorny questions where two reasonable people might disagree.

That's why the law requires platforms to preemptively take down media if someone complains. It's because copyright, in the US and most of the world, is actually impossible to determine for private parties and minor works. You need a court and two sets of lawyers to figure out who actually did what. As the article says:

> The status of RODIK and the ownership of its rights are currently unclear. This makes it likely that Cookie’s Bustle is an “orphan work”, a copyrighted work where the owner is either unknown or cannot be located.

Copyright reforms requiring registration could fix this, but I don't think things are going to be calm enough to allow it for decades. I get that it's trendy to complain about big companies getting this wrong, but it's stupid to blame them for trying to survive under the current rules.

[chrismorgan]

I do speak more broadly than this specific case. But even in this case: many of the notices here might have been deemed not-obviously-invalid, and thus DMCA process should be followed, but also (as described) many of them are manifestly fair use or in some cases not even infringing: and such notices can reasonably be rejected.

Don’t think these platforms obey every notice that meets the nominal requirement of the statute; they definitely do ignore some because they’re obviously nonsense. I suggest they should do this more. Yes, they theoretically open themselves up to liability in doing so, so I do expect them to err on the side of the claimer if there is any reasonable doubt.

But instead, they side with rights-claimers (who may or may not be rights-holders) structurally. They make takedown systems that go well beyond what is legally required, and then don’t police them, so that they invariably become vessels of abuse.

Here is a detailed example from uploading church services with old hymns in 2020, and YouTube’s Content ID system actively facilitating copyright fraud: https://news.ycombinator.com/item?id=27004892. I say YouTube should be required to cut off such transparent copyright fraud, when it is pointed out; and that if they’re not willing to do so, their platform should be shut down.

[WorldMaker]

The core ask in this article (and elsewhere) is not for large companies to prove or disprove copyright claims themselves without the help of the court systems, the primary ask is that companies better support even basic Fair Use checks/reviews before takedowns. A lot of "Fair Use" qualifications are easy: is it academic (an essay, perhaps), is it transformative (a video about a game and not a copy of the game, maybe), how much of the copyright was infringed (is it a short clip in a larger study, as an example).

The big companies don't want to do this basic due diligence because today at least it requires human labor, even if that human labor is "do a quick glance and check a couple boxes".

The article even points out that US laws say that things taken down for copyright infringement but are in fact Fair Use should be able to claim damages. In theory a class action lawsuit of video essayists could make a real strong case in direct, estimated demonitization losses due to spurious copyright takedown notices YouTube acted upon automatically without any Fair Use checks. I can't imagine the stress of being involved in a case like that in practice, which is probably why there isn't enough people begging to be in a class action lawsuit like that.

[SR2Z]

The way it works now is that the accuser has a few days to submit proof that they're suing to defend their rights, or else the platform can reinstate the content. In practice this would (and does) lead to a game of whack-a-mole between large rightsholders who have to pay money for lawyers and uploaders, who don't pay anything but an internet bill. Obviously, this doesn't fly in court and platforms have to go out of their way to ensure that they aren't profiting from mass piracy. It doesn't help that the aforementioned lawyers are always eager to go after a juicy, solvent target instead of Some Dude in Ohio with fiber and a lot of free time.

What do you do when your basic fair use check turns out to not be so basic after all? What happens if a video starts as academic but later turns out to be part of a commercial operation? Is the platform indemnified because it was "obvious?"

You're also forgetting that the platforms do not want to take down content. YouTube at least does a few basic checks automatically and makes heavy use of human reviews. I'm sure a few people would benefit if they quadrupled their spending on copyright review, but it's crazy to think that it makes sense for them to do this.

[chrismorgan]

That’s how it works under DMCA. But some of the largest platforms go ridiculously further.

The way it works now on YouTube is that you get a copyright claim which they probably won’t tell you about, but will just steal your money, or a copyright strike if they want to actively take it down. If you contest it, the rights-claimer gets to decide your fate: if they ignore you for a month, or if they decide your counterclaim is okay, you’re fine; but if they decide to press the other button, your entire account is at least ⅓ of the way to being blocked. And there is no recourse! YouTube refuses to adjudicate. This system is insanity.

[kiba]

Copyright is a special privilege that usually monopolizes certain information.

It has been recognized by early Georgists as monopolistic and thus problematic.

That's why you see rent seeking behaviors and abuses. It's not new.

[SR2Z]

"Usually?"

The point of copyright is that it's a legitimate form of ownership. It's an acknowledgement that someone who creates a work should own it.

It's not any more monopolistic than owning a car, for example.

[WorldMaker]

A copyright isn't owning a car, a copyright is more akin being the only person legally allowed to manufacture cars. (This isn't hyperbole, a patent on cars is control of the very concept of cars; copyright and patents are more similar than dissimilar.)

That's why it was supposed to be a limited right with a clear and simple expiration. No one should own the concept of a car forever, eventually you want other people to be able to manufacture cars.

That's why it is a monopoly right.

[SR2Z]

A patent lasts 20 years. A copyright lasts for the life of the author plus 70 years (at least in the US, thanks to Disney's lobbying efforts).

The term of a copyright is too long, I agree, but it is a limited right to control a creative asset with a clear and simple expiration.

[WorldMaker]

Right, it is a limited right because it builds an artificial monopoly. Copyrights were intended to have a similar life cycle to the patent. It lasts for a few years, possibly with a single extension if you can prove certain things about how you are using it (that you are actually using it, not just squatting on it to prevent other people from working with it).

It is a bit broken that the term limits are so different today.

> a clear and simple expiration

> life of the author

> (at least in the US

I think you included several reasons it is not clear and simple. Life of the author is real hard to define and gets shifted by "work for hire" rules, especially because so many things subject to copyright beyond books don't/cannot have a single author.

On top of that, different countries have different definitions. The Berne Convention muddies the waters that "the strictest country's definition wins" but also provides carve outs for "when in my own country I only need to worry about my own country's rules" some of the time.

Different countries have different orphaned works laws, though the majority do not today believe copyright expires on orphaned works it just gets "lost" who owns the copyright. Most countries have "copyright is automatic" laws (and the Berne Convention supports that) and "copyright is assumed and must be disproven" laws (which again the Berne Convention supports). All three of these things make the question of "is this under copyright and by who?" far from clear. (As the article here goes at great length to provide just one example of such confusion and opaque expiration information.)

The world's copyright systems lost "clear and simple expiration" decades ago.

[BrenBarn]

Amen to that. Larger companies should face more requirements than small ones. And if they can't meet them, great! That just means they can't be large. Companies should be smaller on average than they currently are.

[SR2Z]

There is nothing that large companies want more than a straightforward, procedural list of requirements on how to comply with copyright law.

(ok maybe there are several thousand things, but you get what I mean)