> The small number of humans available to review contested claims means that people who fall afoul of machine error, sloppiness and criminal mischief are often unable to get a fair hearing or justice.
This is possibly one of the key points in this report. Google is synonymous with their blackbox/magic "AI".
These systems become remarkably interesting to game (their ad platform has been subject to fraud, but on a different layer). I wonder how many levels of "AI" they think they can automate before they realize that it's probably not possible.
For some good news though, there was a story a little while ago for some educational platform that noped out of YouTube and migrated to self-hosted.
We thought it would take some sort of super AI to take over the world just make arbitrary / unfair at face value / decisions that can't be appealed or undone.... naw we just needed companies like Google.
I think for Google at this point we can be sure, there simply isn't a decision making process. Nobody is deciding things, nobody is responding, it's just a black hole where a process leads to and ends.
I'd say that Google is synonymous with hubris. Their unexamined but faulty trust in their ability to create automated systems that don't require manual intervention or customer service is one main application of it.
OP is just wrong here. Copyright strikes don't go through content ID, and anyway nobody at YouTube is responsible or should be (per DMCA law) to decide what's infringement. If there's a dispute it can go to court.
YouTube is personally liable if they reject valid DMCA notices. The only way to keep the safe harbor and guarantee immunity is to follow the DMCA process. You're suggesting they voluntarily give up protections to hurt bad actors.
They are not required by DMCA to nuke you after 3 strikes, and they are allowed to keep the content if you file a counter-notice. Neither really works with YouTube. This IS on Google.
The article disproves your opinion. They are, with no punishment to bad actors, i.e. there's no 3 strike for absolutely false copyright claims, stealing money from content creators and violating the copyrights of content creators. If anything they should lose their safe harbor for facilitating copyright infringement due to their 3 strikes system that has no way to circumvent the false claims. Even if you file a counter notice, the claims are often upheld unless there's a massive outcry. Bad publicity should not be the only recourse for bad actors.
>They are, with no punishment to bad actors, i.e. there's no 3 strike for absolutely false copyright claims
This would, in fact, cause them to lose their safe harbor under DMCA. There is no provision to simply ignore claims.
>stealing money from content creators and violating the copyrights of content creators.
Who's copyright is being violated? What money is being stolen?
>they should lose their safe harbor for facilitating copyright infringement due to their 3 strikes system that has no way to circumvent the false claims.
I strongly recommend that you actually read the DMCA.
The safe harbor is given to the service provider for accepting DMCA notices, and protects them from suits from the complainant. There's a separate safe harbor against claims from the uploader, and they get that by accepting counter-notices. But there's no obligation to accept counternotices, since they can delete videos at any time per their terms. So there's no relevant safe harbor they would lose here.
You've provided no source for the claim that they often reject valid counternotices. I don't believe this is the case.
There is no AI to resolve contested copyright claims. It just sits in Google's queue for weeks because Google refuses to hire enough employees to resolve issues in a timely manner. "Computers taking jobs" would arguably be better because then somebody would do the job, instead of the job just not being done at all.
ContentID is their AI. And of course it doesn't resolve anything - "AI" being opaque and unpredictable by definition - but parent wasn't saying that computers would do a better job than humans, just that they'd be a lot more convenient.
I hate to cheer on a crime, but I actually think this is a positive development. Fake copyright takedown claims are already used to chill speech, but it flies under the radar so the process stays broken. If this type of ransom takes off, YouTube will be forced to actually address false takedowns.
Or not, they will just protect the bigger channels that would get articles written about them and hurt their ad revenue. Everyone else though will just get the classic google cold shoulder.
Most probably yes. Google hasn't been friendly to his content makers for a while.
They were nice before they had the monopoly. Now they don't care.
That's why I really hate all those speeches about how it would be great if everybody used chrome.
A lot of thing the privacy savy tech community predicted about Google 10 years ago arrived exactly as expected. And the reason is not that they are seers seing the future, but because it's always the same story with entities growing and getting power.
YouTube has always had the monopoly - not on video distribution, as many mistakenly assume, but on monetization.
Twitch has recently encroached on that front, and therefore YouTube Livestream was made.
The real reason YouTube doesn't care about content creators is that they are trying to woo more mainstream media onto their platform. With the decline of ad revenue since the ad-pocalypse of 2015-ish, they are trying to change their platform to look like traditional media where big advertisers would pay for spots (the likes of Coke and Nike).
Not on a TV for example. Even Pi Hole doesn't cut it since YouTube/Google do their best to use random URLs that are close to impossible to block entirely.
I actually connect everything through Algo and use that ad-blocking.
This didn't work because the ad was just merely another youtube video. One I've watched before, even. They're definitely trying to boost the view count of that video (it's the first 23ish minutes of that new Gundam movie).
>They were nice before they had the monopoly. Now they don't care.
Google was never nice but it didn't matter before the monopoly. They would literally ban and shutdown all accounts based on automated 'strikes' but most users had other emails/search engines/maps/whatever.
Now that Google has firmly entrenched itself, it's impossible to jump ship. Look at the other thread about Google Analytics and integration with Google Ads; its quite a difficult task to switch from GA
Google had a stellar reputation during the first decade of their existence. Their were the paragon of all the geeks around the world: promoting freedom and openess, hacking and innovating.
And when they bought youtube, they were very, very nice with content creators first. Bot were not banned. No detecting of your white noise video as a copyright material.
The danger with you narative is to forget the evolution from underdog to leader. Because people need to learn this happens all the time, and they need to learn what it looks like.
It'll hurt the class divide even more. People without the cultural knowledge that you need to escalate to Twitter/PR or to Google's legal department (eg, certified letter to their process server explaining your possible legal case against the extortionist and demanding they preserve evidence) will be even further out of luck.
Cultural knowledge, or just working as a mainstream journo as jack has pretty much confirmed that they are the lifeblood in his eyes so they can't be banned.
You only have to look how journos were not even being suspended for calling for the complete destruction of the maga kids.
While if your a conservative, make one offensive joke about looks and your permabanned.
That's what happens when there are no consequences for bad actors. Is anyone surprised? A legit copyright claim can take down a copyright violator and "restore" the copyright holder's control of the copyright just as easily as a bad actor can take down a YouTube publisher with lies based on the DMCA. The law equally protects both actors. Greed makes lawmakers look stupid when in fact they are just vile when creating their laws. Then Google comes in, looks at the same problem, and concludes that the law is enough because their content producers are not customers (they are the product, you guessed it) and therefore should be exploited rather than protected. Abusers of this law will simply not be punished. Undoubtedly this sentiment is shared across a variety of major platforms whose only incentive is to pretend like the systems they set up are effective at enforcing a law that's simply unenforceable. And so the charade continues with people's livelihood and creative works as the cannon fodder in this stupid war of greed. Someone should get together and start filing such claims against content the big media corps provide. I guarantee you Google fixes this problem in hours.
I think youtube is actually one of the best demonstrations of why that attitude is incorrect.
Obviously it's better if you haven't trusted 100% of your business to a platform you don't control. But for some businesses (such as many youtubers) the choice is either do that or don't have a business (since that's the only place with a large enough audience). In such a case it can make sense to do so.
It depends on the youtuber of course, a few of them have an audience that legitimately allows for a plan b.
For most of them there isn't. They might have seperate portals which they earn money from, but the business still goes poof fast if they get banned from youtube. Without publishing on youtube they won't reach enough of an audience to have a sustainable business.
Twitch alleviates this somewhat for some youtubers. But we don't have to go far back in time to consider an era before that existed.
Not sure I agree with this. Beyond a certain level, subscriptions to a given creator are to that creator, not via the platform. It's true that a lot of consumers won't take the effort to follow the brand off platform, but that's not always the case. I think the recent development of the influencer marketing industry as a whole is proof of this.
Of course, how you ever get remotely famous or established in the first place if the content you create is remotely vulnerable to false copyright claims is another question entirely...
Patreon and similar platforms make your income stream independent of YouTube and give you a second channel of communication to your most dedicated fans that also allows new content even while your YouTube channel is down.
no, but you get some kind of insurance when youtube screws you over, and also even if you get the eyeballs there, you don't get the money there, which is still a bit of protection.
There was this story recently about Youtube maker Naomi Wu who first got had her Adsense blocked on Youtube, moved to Patreon, got blocked there, moved to subscribestar and it got nuked. She has her own site but because she is a Chinese resident it makes, according to her, payment services scared and difficult to work with. In the end she returned to doing programming work for hire to pay the bills, and making/video creation as a hobby, even if she has a large enough community that could support her creative work.
It is a hard problem to solve. I would not know what for example Jim Sterling would do if he got blocked on patreon, as all his videos has multiple false copyright claims to them. I know Linus tech tips (an other large youtube creator) has created their own platform as a massive Plan B, but that relies on payment services which could at any time destroy that option.
I'm surprised she didn't use bitcoin or some other type of cryptocurrency. This is the exact type of scenario they are designed to address. If her fans are willing to use third-party services like patreon to support her, I can't imagine getting their hands on some bitcoin being too much of a barrier.
Youtube being sued by Hollywood company with big pockets vs angry uploaders of personally-generated content who don't have much money. At the moment, it makes business sense for Youtube to ignore angry uploaders.
YouTube's "safe harbour" takedown system is a bit stricter than what the law requires - it doesn't require an actual DMCA takedown (which, when falsifies, results in a fine of $800), but simply using YouTube's internal system (which has no such falsity fine).
If you're referring to Content ID, you're correct that it's not a DMCA takedown, but if disputed and appealed the claimant must file a DMCA to keep the claim active.
Hank Green posted a video just yesterday that gives a lot of good context about the YouTube ContentID system, its challenges, and the broader impacts of copyright, creativity, and culture.
I have a youtube channel where I post videos of dance competitions. Because it involves music, I frequently get copyright claims from whatever rights group that wants to take my ad revenue (which is 0 anyway because I disable monetisation on principle).
When you get a claim, there is an option to appeal it, all the way to a formal counter-notification which will remove your strike and require the claimant to institute court proceedings against you.
I have even used that a few times, years ago when I still got fired up by this. Now, I just appeal it once on fair use grounds and if it doesn’t get dropped I move on with my life.
Has something changed that doesn’t allow this particular channel owner to challenge the claim? (Genuinely interested as a channel owner myself.)
> which is 0 anyway because I disable monetisation on principle
I also disable monetization on principle, but I think if a video is claimed by a copyright holder, then YouTube will add adverts in it, even if you disabled it.
As to your other point, a claim is different from a strike from what I understand. A strike is a more serious offense that may lead to blocking the account entirely. A strike can come if you dispute a claim and the claimant is considered to be in his right.
I know the distinction between a claim and a strike - I have direct experience of both. ;)
You can appeal a strike and it will be removed unless the claimant provides documents that they have instituted proceedings within a fixed period of time.
You have to provide your contact details if you appeal a strike.
From what I understand, YouTube doesn't review challenges to claims, the actual claimer, the blackmailer will review the claim. And when the claimer rejects a claim the youtuber will get a strike. 3 Strikes and you are banned.
But from copyright point of view, do you actually have right to post those videos with audio? I'm not convinced that it would fall under fair use, so wouldn't you need explicit license to publish the videos?
Under UK law it may fall under the protection of coverage of a current event (this being international competitions organised by a OIC member body) and also the grounds of it being partial recording of a track and low quality (to establish no commercial loss).
It is a very shaky argument, hence I give it one try and move on (it does work about 20% of the time).
I believe the difference is that you “move on with your life” which I assume means you take down the video. If you had written/produced the song and were trying to monetize it, it would be a different story. There was an article on HN recently where someone remixed a song they found on youtube and then used the remix to claim they had the copyright. YT assigned the rights of the ad revenue to the claimant despite the fact that the rightful author denied all allegations.
Your belief is somewhat misplaced. ;) I just mean I leave the video with the claimant monetising it. It is very rare that they would chose to block the audio (except for some that are blocked only in Germany which has really insane copyright laws). So I just leave them as they are, without taking them down.
But you are right in that because I do it for love rather than money, I am in a much more comfortable position.
As I mentioned in the comments below, there is the option to appeal it to a point they would have to sue you. If you are sure that you own the copyright that is an option. This may be of limited comfort if you happen to live in a jurisdiction with a f’d-up legal system (e.g. the US) but that’s not Google’s fault.
That said, there is apparently some sort of underhand agreement between Google and some major music publishers that Google will just take down the content privately with no appeal. That sucks, but it’s not what is happening here, I don’t think.
We desperately need another major video streaming site. Google has been consistently acting in an atrocious manner for users and content creators, with negligible cost due to their functional monopoly - even massively popular creators are able to recover their videos only in the rare situation of widespread media attention [1].
I've decided to use it to host some artsy stuff as well. I don't know what will happen when I reach the 5GB total storage limit though. I'm not keen on paying a monthly fee for the rest of my life so my videos, including the older ones, stay available...
People have been treating pornhub as a site to put random non-pornography content at least since a couple years ago according to my friend, I haven't encountered it though
Make the financial consequences of bogus copyright claims expensive and enforcable.
Youtube needs to make it impossible for "VengefulFlame" to register a copyright claim, without a real-world identity attached to it and a ironclad agreement that they 100% have the right to enforce this specific copyright claim, under agreed up front financial penalties if they are wrong (even if they're accidentally or algorithmically wrong, they still owe the penalty). Youtube should pay (or perhaps 70:30 split?) those penalty payments with the content creator who was fraudulently or accidentally accused, if paid out via Youtube T&Cs and contractual arrangements, and Youtube should provide all required evidence and documentation (including verified real work personal identities and their digitally signed assurences to Youtube of copyright ownership) to the victim in a super-easy "boilerplate" form, so that it'd be a slam dunk legal case that any lawyer with an associate with an hour or two spare would happily take on on a no-win no-fee basis.
There's probably some cross jurisdiction legal difficulties to address, but I reckon YT should step up to actively take abusers of copyright claims to court in their own jurisdictions on behalf of their victim content creators. If you live in a jurisdiction without strong civil copyright law which includes penalties for fraudulent claims, maybe you should pay a bond before being allowed to file any copyright claims?
The underlying problem here is that Google has no capacity to adjudicate copyright claims. How the heck are they even supposed to know who the copyright holder is? The troll claims it's theirs, the user claims it's theirs, they have no basis on which to believe one or the other.
In theory you can register with the copyright office, but it isn't required and >99% of YouTube users won't have registered anything, so then what does that prove?
This before you even consider fair use, which is hard enough for judges after a year long trial, much less something they're expected to decide thousands of times every day.
So you have the claimants post bond. Then what? You still need a way to know which claims are false, which they still haven't got.
They're making decisions based on no information. They're trying to play a role they're not equipped to play. Google is not the court system.
For starters, actually enforce the required elements of a DMCA notification:
(i) a legal signature
(ii) identifying the infringed content
(iii) identifying the infringing content
(iv) proper contact information
(v) statement of good faith belief the content is infringing
and (vi) statement the information is accurate and statement under penalty of perjury of authorization.
Last I saw every pseudonymous troll can make claims, not fulfilling (i) at the very least. I remember I used to click some of those chilling-effects/lumendatabase links google displays when they removed some things from search, and I remember I found it rather curious that a lot of these usually where incomplete notices, usually failing (i) (ii) (v) and (vi) (and they only got (iv) right because they emailed and every email has a From:).
"How the heck are they even supposed to know who the copyright holder is? The troll claims it's theirs, the user claims it's theirs, they have no basis on which to believe one or the other."
The DMCA has a procedure for that:
1) Copyright "owner" claims content with google
2) "Uploader" files a counter claim with google
3) At which point google has to put it back after 10 days if the "owner" doesn't notify google they filed an action against the "uploader" in a court.
Yes, this is still far from optimal, for all parties involved, however, it's still a lot better than what google does with ContentID.
I'm not saying Google should adjudicate on the copyright ownership. (In fact they probably should explicitly avoid doing so, as you point out.)
What they should, in my opinion do, is ensure that if a claim is made, and it's then objected to - that the claimant has a realistic risk of direct financial penalty for doing so fraudulently. Google should demand all relevant documentation for both parties before accepting or acting on a disputed claim, so that either or both of them have a lawyer/court ready package of the specific claims, counterclaims, and identities of everybody involved.
I also think it would be smart and desirable for Google to have an internal mediation system with a penalty enforceable by T&Cs for Google Accounts. If, for example, Sony Music's lawyers incorrectly claim copyright on a content-creators piano performance of a public domain work (which has happened more than once), and the performer objects, it would be nice if Sony had the option to say "Ooops, sorry, charge our Google account the $500 erroneous claim fee, and pay it to the content-creator" to avoid having the content-creator's lawyer slam dunk them in court with the documentation Google provides as above.
There should also be the same sort of "three strike and you're excluded from all google services forever" nuclear option, that applies equally to copyright claimants as it does to content creators, and also applies equally to Sony Music and VengefulFlame's legal identity. Sony need 100% to be at-risk of having _their_ content deleted from Youtube and their Google accounts permanently revoked for sending fraudulent copyright claims, just the same as the rest of us risk having our Google accounts shut down for illegally uploading Sony's copyright works.
They could fall back to the legally required process (DMCA takedowns). There is no law requiring Google to take down channels because of reports on a few videos, or requiring Google to keep content down after the uploader has re-affirmed it is not breaking copyright law (called a DMCA counter-notice).
That's not even remotely what we're talking about here. Anyway, this has nothing to do with DMCA but with Google's Content ID system which is much worse than DMCA.
No, that's not how it happened. ContentID is a proactive system that goes well beyond what the DMCA requires and skirts around the DMCA penalties for false claims.
The DMCA requires big parties like YouTube to create systems like ContentID and the copyright strikes to remain shielded from copyright infringement claims. You can be sure Google isn’t running these systems because they want to.
Not exactly. As long as YT acts the way the DMCA specifies (receive takedown notice, take down content, notify account owner), YT maintains safe harbor protection.
The problem is that Big Media doesn’t like the idea of safe harbor. It also doesn’t like the potential to have their unlicensed content available while a court decides (they file a DMCA takedown, but the account owner claims “fair use” and YT puts it back up - now it’s online until a court orders it offline.) Hell, Big Media hates the idea of fair use.
ContentID came about because Google decided it would be cheaper to just allow Big Media more direct access than to keep defending themselves in spurious lawsuits designed to make example cases and have parts of the DMCA nullified.
They aren't spurious lawsuits. Google is pirating content and selling ads on it.
Everyone knows that. YouTube built itself on stolen content. Any of us could go on there right now and watch any number of pirated movies, songs, TV shows. .. you name it.
The issue seems to be that it's nearly impossible to find human reviewers and a lot of the system is designed around lying to you about it. Someone made a video about a year ago about this, but even if you request manual reviews and reply to their emails, you consistently just get emails from YouTube that are automated bots, disguised to seem like humans: the emails have a name in them, the reply period is randomised, etc.
Even people with millions of subscribers are not immune to this, in fact they often have to hope their fans can cause a big enough mess on Twitter in large enough numbers for YouTube to notice and suddenly all those "manually confirmed" copyright flags disappear.
Even bigger of an insult to injury is that there are clearly corporate YouTube accounts that are above this system, the accounts of Kimmel, Trevor Noah, etc are immune to any of these.
And it's not just copyright, but monetisation, too.
Currently Youtube's system encourages this behavior by offering the possibility of profit[1] without any risk. There are no consequences for making false claims (even when the claim is very obviously fraudulent). There isn't even any risk of Google banning the account, while creators are given "strikes"[3] against their account. Google is incentivizing fraud wile driving away a lot of talented creators.
The only way this can be fixed is by reversing the incentives. There has to be consequences for making false claims. There must be some amount of risk when making a claim. Also, Google could at least improve the situation easily by e.g. holding a video's revenue in escrow until the claim is fully resolved instead of giving the money t to the claimant immediately (I call that "theft").
[1] YT allows successful claims to force monetization on a video and take the resulting revenue. There are numerous examples of e.g. musicians having their own music claimed, or people losing the income from multi-hour educational videos[2] because ContentID matched 13 seconds of random footage.
[3] On the 3rd strike, Google automatically delete their entire YT channel and a the "offender" is permanently banned of from all Google services.
TL;DR - video [2] has an outstanding explanation of YT's takedown system from a creator's perspective (YT took down His 4 hour thematic analysis of Made In Abyss).
If that's the solution then do be it. Hire more people. "We can't properly handle copyright violations because our algorithms aren't good enough" is no excuse. they are selling ads on copyrighted material. It's illegal.
It's the same problem. Google can say we have a content ID system to handle copyright violations. See how good it is?
Behind the smoke and mirrors they can make money on pirated material.
All the while throwing their hands up in the air. When all they have to do is hire a human being to evaluate the facts. People know what copyright violations are. Pay people to make the call.
It's as simple as that. Hire people to do the job and it will be done. Computers don't do it right.
Do you want to hire people to evaluate flags and disputes, or to evaluate all videos? If just flags and disputes then there will be false negatives on pirated material that no one has flagged yet. If all videos, that would be very expensive because 400 hours are uploaded every minute[1]. And it still would have false negatives, because how is the human reviewer supposed to know that the song played at 3 minutes into some video is a copyright violation or not? The song could have been an original composition, or it could have been licensed by the musician to the youtuber. You can't expect a human to have knowledge of every piece of copyrighted content in existence to be able to know if the video matches any of those pieces of content.
An OSP must "not receive a financial benefit directly attributable to the infringing activity" to qualify for § 512(c) protection. However, it is not always easy to determine what qualifies as a direct financial benefit under the statute.
One example of an OSP that did receive a direct financial benefit from infringing activity was Napster. In A&M Records, Inc. v. Napster, Inc.,[10] the court held that copyrighted material on Napster's system created a "draw" for customers which resulted in a direct financial benefit because Napster's future revenue was directly dependent on increases in user-base.
How is it possible to review all of it? How can a human know whether a song is infringing someone else's rights or is an original work by the uploader?
You're making a valid point in general, but do note that in this particular issue the problem is that they are actually overzealously removing videos with potential copyright infringement.
I might suggest that is their intent. They can point to the articles exclaiming that it quickly responds to claims as a cover for all the actual pirated content they earn ad dollars on.
I'm a little tired of all this creators complaining how broken is Youtube (and other platforms) and still investing lot of their time and effort in building channels, accounts and entire businesses relying on these platforms.
For them moving to another platform doesn't make sense because they wouldn't get the same reach/make enough to justify the time and effort put into their channel in the first place. So it's either they are on YouTube and complain of its faults, or they aren't anywhere and you don't hear about them.
Until the platforms push exclusivity contracts... I do remember a twitch streamer being happy about his contract being old enough to allow him streaming on youtube while temporarily being banned on twitch. Sure, you can fault the individual for signing something like that, but an effective monopoly will hardly leave you an other choice.
This is possibly one of the key points in this report. Google is synonymous with their blackbox/magic "AI".
These systems become remarkably interesting to game (their ad platform has been subject to fraud, but on a different layer). I wonder how many levels of "AI" they think they can automate before they realize that it's probably not possible.
For some good news though, there was a story a little while ago for some educational platform that noped out of YouTube and migrated to self-hosted.