YouTube's Content ID System Flags, Demonetizes Video Of Cat Purring
from the purrrrfect dept
YouTube's Content ID automated copyright system sucks. There, I said it. Any review of the different posts we've done specifically on the topic of Content ID can only leave you with one impression: the system doesn't work. Not that it never works, of course, but when you build a system that is designed specifically to allow 3rd parties to take down speech content, that system had damned well better not be wide the hell open for abuse or laughable errors. Well, guess what? You've got your music labels getting works taken down that were specifically designed not to not be infringing, news organizations managed to claim their own live streams as copyright infringing, and music labels being able to demonetize videos of a guy singing public domain Christmas carols. It's all very stupid, very much the tip of the iceberg, and very much an indication that Content ID, in its current state, is broken.
What's that, you say? You need more? Fine, a guy uploaded videos of his cat purring and those got claimed by two different labels as infringing on their copyrights.
YouTube's automated takedown tool is known for its flaws, but this week it crossed a line by attacking a purring cat. According to YouTube's Content-ID system both EMI Publishing and PRS own the rights to a 12 second purring loop. Last March, YouTube user Digihaven uploaded one hour of video loops featuring his cat Phantom, purring, as cats do. The video didn’t go viral but appealed to a niche public, and more recently also two major music publishers.
Nearly a year after the video was posted Digihaven was informed by YouTube that Phantom is “pirate” purring. Apparently, part of the 12 second loop belongs to EMI Music Publishing and PRS.
Yes, this is sort of funny, but only after you've encountered so many Content ID problems just like this that you become dead inside, like me. I've made statements like this before, but I'll repeat it again: your automated copyright system doesn't have to be perfect, but if your system is so flawed that a 12 second video of a cat purring can be flagged by multiple music labels then your system sucks so badly that you need to completely start from scratch on a new one.
Now, I've done some haphazard searching for songs entitled "Focus" by artists on the EMI label and, frankly, I gave up. There are a ton of search results for songs that include that word in their titles. That being said, I'm fairly certain that EMI doesn't have a band with a song that is 12 seconds of a cat purring or, if that in fact is a thing, that such a video would be copyrightable. In other words, either way, EMI and PRS should not be monetizing Phantom the Cat's musical stylings.
“I’m sure EMI/PRS made Phantom a sad kitty. It seems like companies such as EMI are pirating ads on people’s legit videos, so I’m wondering if they apologize to, or reimburse people for those false claims,” he tells TF.
Hoping to clear his cat’s name Digihaven decided to file a dispute. This was partially successful, as EMI lifted its claim shortly before publication of this article.
Which, absurdly, means now Phantom just has to figure out what PRS' problem is.
Or, hey, maybe we could all just admit together that Content ID in its current form doesn't work and should be done away with in some organized and planned fashioned. Replace it with a better automated system. Replace it with more humans doing moderation. Admit that content moderation at scale is completely impossible and stop trying.
Anything would be better than living under a automated system we all know sucks.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: cats, contentid, copyright
Companies: emi, google, prs, youtube
Reader Comments
Subscribe: RSS
View by: Time | Thread
Well the industry told us that pirates had made music worthless even as they paid 4 billion for the dead carcass of EMI back catalog.
This obviously was their plan all along, to just have the zombie label claim everything in existence to claw back the 4 billion they wasted on worthless music.
Or they are a bunch of incompetent assholes because there is no punishment for being wrong but huge penalties for not jumping fast enough to act & huge legal bills for those wrongfully accused by zombie label to have a judge declare that the cat purring isn't one of zombie labels content & not make them pay the bills for the person they wrongly accused.
[ link to this | view in chronology ]
Re:
There's two groups that both need to suffer penalties for abuse:
The rights-holders who file claims without doing their due diligence to determine whether they really hold the rights to the material or not. There needs to be statutory damages defined for filing a takedown request for material the filer doesn't actually have rights to, and those damages need to be non-discretionary.
The hosting providers who go along with the rights-holders by default. There need to be statutory damages defined for repeatedly accepting baseless claims, and for continuing to act on claims from rights-holders with a record of baseless or abusive claims. I suggest a three-strikes rule: three baseless claims from a rights-holder and they're subject to statutory damages for every baseless claim, three baseless claims from the same rights-holder against a hosting provider and the hosting provider is subject to the same statutory damages if they continue to accept claims from that rights-holder.
[ link to this | view in chronology ]
Re: Re:
"There's two groups that both need to suffer penalties for abuse:"
No, really just the one. The hosting providers are helpless since if a rights-holder files a complaint it's up to the hosting provider to either remove the content filed against or shoulder all the legal risk.
This is the problem with the DMCA effectively reversing burden of proof. The one filing a complaint is always assumed to be in good faith - which they never are - meaning the safe harbor provisions are all which stands between the hosting provider and staggering fines.
I've said it many times before and I'll say it again; Copyright is beyond salvaging. The harm it causes far outweigh any potential benefits.
[ link to this | view in chronology ]
Re: Re: Re:
Yeah, DMCA is an issue.
However, ContentID goes beyond what DMCA requires. YouTube should be held accountable for it.
[ link to this | view in chronology ]
Re: Re: Re: Re:
"However, ContentID goes beyond what DMCA requires. YouTube should be held accountable for it."
Pressure from the industry. ContentID is part of Youtubes "retain safe harbor" card which lets them minimize the takedowns - and subsequent contesting of those takedowns - which lets them keep the amount of people tasked to adjudicate within halfway reasonable limits.
Remember, every industry action is a consequence of a cost-benefit analysis. ContentID exists because it saves Youtube money and overhead in a paradigm where copyright legislation is outright insane.
[ link to this | view in chronology ]
Re: Re: Re:
Copyright has a purpose, If you create something someone cant copy it on the spot, sell it, and claim its his. Movies with 200 Million budgets, books, and the newest music all have problems with out some level of copyright.
Copyrights problem is that the memory effect is MUCH longer then human memory. We've got 100 years of history that someone who makes something today can sound like something 40 years ago and the grandkids can make a claim on it now just because it sounds the same. Never mind that the harmony came from something 80 years ago and those great grand kids will make a claim on BOTH people.
Copyright needs to be much shorter. Short enough that reasonably abandoned works can get new life, perhaps give copyright a "Still relevant in the marketplace" exemption or such if the original owner is still actively using it. But we need a way to clear out the trash much faster then 100 + years.
[ link to this | view in chronology ]
Re: Re: Re: Re:
Agreed. The 28+optional extra 28 system had its problems, but it was far, far better (light decades, even) than the current system we have. © was opt-in, not opt-out. There was an option for renewal so if something didn't make money the first term around it would enter the public domain. Also, © didn't take up an entire person's life and seventy years after that, 56 years max is, IMHO, extremely acceptable seeing as what could have entered the public domain (even stuff from the 1990's that didn't succeed!).
It just sucks that greed from the landlords meant that it took so long for us to have nice things.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re:
No way. 11 years, as originally conceived, with one option for renewal. If you haven't made (enough) money in 22 years, then so sad, too bad.
No works are made in a vacuum: we all benefit from the commons.
Star Wars itself was a rehashing of the cowboy western + hero's journey tropes, nothing new except the details.
Every single movie from Disney's Golden Age was stolen er, borrowed from Bros. Grimm (who themselves were thieves)
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re:
My point is that Steamboat Willie was created when 56 years was the maximum term but now © is at an ungodly scope and at an ungodly length (though I am extremely grateful it's at least expiring and the Music Modernization Act means I don't have to pay extra for cover songs of ©'d music unless I upload them onto bandcamp).
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re:
"No way. 11 years, as originally conceived, with one option for renewal. If you haven't made (enough) money in 22 years, then so sad, too bad."
I'm not sure there's a perfect solution, but I've always been in favour of the idea that there should be an initial term similar to the original term of copyright. This allows for smaller creators who depend on word of mouth and other types of marketing to benefit (there's plenty of examples of slow burn successes). Past that, renewals are allowed, but they must be made manually and by the original copyright holder up until their death. After that, the copyright can be sold or transferred but does not come with renewal rights - this allows their estate to benefit for a time, but does not allow infinite locking up of culture by corporations.
Not a perfect system, but that seems to satisfy the requirement for creators to benefit if they decide they need to so, while freeing orphaned works and not allowing long term corporate control of the most profitable works.
"Star Wars itself was a rehashing of the cowboy western + hero's journey tropes, nothing new except the details."
Now now, be fair. Lucas also cribbed a lot from samurai movies, as well as world war footage, not to mention the sci-fi serials that inspired his stories after he failed to get his Flash Gordon project off the ground.
Lucas and his team did incredible work that's inspired generations, but you have to be very blind to not see all the influences he proudly utilised.
[ link to this | view in chronology ]
Re: Re: Re: Re:
That is not a copyright issue, but rather an attribution and plagiarism issue. Protect the right to attribution, and people know who created a work, and who to support if they want more work from the same creator.
Copyright protects the business interests of the gatekeeper middlemen, and the expansion they are driving in copyright and its protections has everything to do with reducing competition, and nothing to do with rewarding the actual creators. Piracy is only the drum they keep beating to try to gain what they really want, control what what works get published, along with most of the income from those works. They would destroy the means of self publishing if they could, and will cripple it as much as possible by forcing everyone to use filters that over block works that are self published.
[ link to this | view in chronology ]
Re: Re: Re: Re:
"Copyright has a purpose". So what? Evil has purpose too. Supporting or creating a system that regularly censor and oppress the mass so 1%ers can make outrageous amount of money at the expense of the mass is a purpose I don't care for. Forget corporate welfare. How about caring more for public welfare? Surely this is a more noble purpose.
What do propping up highly ineffective business models has to do with the greater society good? You think supporting ineffective production of Hollywood-style accounting movies with 200 million budget for example at expense of society's freedom and free markets worthwhile enough to justify copyright monopolies and is a good for the greater society? Surely, more and superior movies could be produced for less and available to society for less than monoploy-rent inflated prices with an alternative economic system that encourages efficiency in the market for cultural goods instead? Big budget movies
don't correlate to quality, you know, and this is especially true of Hollywood. This Hollywood style inefficiency and relative lack of quality in relative to cost of production does not speak well of copyright monopolies to me.
Also instead of society wasting wealth by overpaying for their copyrighted, overrated, and overpriced cultural goods to the 1%ers, surely there are more valuable goods that wasted wealth could go instead to produce? The opportunity cost that copyright monopoly rents introduce to economy is should be taken in consideration in any economic argument for copyrights. is copyright monopolies indeed worthwhile given the opportunity cost?
Okay, is a world without digital copyright monopolies so horrible to contemplate? To be sure, it comes with problems but so what? A world with copyright monopolies come with problems as well. what's the difference? which problems and challenges are better to have, that is the question.
[ link to this | view in chronology ]
Re: Re: Re: Re:
"Copyright has a purpose"
None which justifies the drawbacks. You could fit created works under trademark law - make it part of a creator's brand - and solve almost every issue that way. Copyright though, in order to exist, infringes on every persons personal rights of property and comes uncomfortably close to old-style heresy law from medieval times in the way it exercises information control.
"Copyright needs to be much shorter."
It used to be. Once upon a time it was 7 years. The lesson of history is that as long as this fundamentally flawed legislation exists enormous monetary incentive will exist for middleman gatekeepers to lobby for extension. That being the case you literally can not restrict it, reform it, or pare it down with a single hope of keeping it within reason. Copyright will only ever have two settings - abolished or utterly unreasonable.
[ link to this | view in chronology ]
Re: Re:
"The hosting providers who go along with the rights-holders by default"
They don't really have any choice. They don't have any control over the content before publication, so they have to react to what happens before that (unless there's a ContentID or similar match on upload, which is equally full of problems on a different level and is really only an attempt to fend off the huge level of notices received).
Then, the choice is to either obey what should be a DMCA notice that's issued under penalty of perjury from the claimant, or fight each individual request in court. Given that YouTube was at great risk of going under due to Viacom suing them over files they uploaded themselves, and that nobody except YouTube really has the resources to even consider going to court with millions of claims, hopefully you see why this blame would be misplaced.
It would be nice if other ways of reacting were available, but if you're blaming providers for obeying legal takedown notices, your actual problem is with the DMCA, not the people trying to obey the law. At best, you can argue that takedowns have to be verified by humans before a takedown occurs, but this is a huge undertaking that can't be solved just by throwing people at it, as even a tiny margin of error still means thousands of videos blocked by false claims.
[ link to this | view in chronology ]
All gain, no pain
So long as there are no penalties for bogus claims there's no reason not to claim and monetize everything you can get away with. At most they might end up dropping a claim, but with the system so one-sided that a cat purring can be claimed I doubt that there's any penalty whatsoever if they decide that they do in fact have the rights, and right to monetize, whatever they claimed dibs on.
[ link to this | view in chronology ]
Re: All gain, no pain
In theory, John Cage's 4'33 is a copyrighted work. It's also possible to automate takedown notices for what is meant to be a statement under penalty of perjury from the individual making the claim.
If the system is so broken that these things are possible, then until copyright law is fixed to a more sane version, the sad fact is that they can claim whatever the hell they want and whoever gets the legal takedown order needs to decide between the victim and the abuser. Hosts are unlikely to be getting enough significant revenue from the car purring video for them to be able to fight it.
[ link to this | view in chronology ]
That Torrent Freak news post about a purring puss is seven years old.
[ link to this | view in chronology ]
Re:
I remember that post! I also recall more recent stories on videos of birds singing and planes landing getting flagged by contentid, which could probably have served as effective examples that it has not improved
[ link to this | view in chronology ]
the US Copyright office just had a consultation on these systems and - as you'd expect with a Registrar of Copyrights that was the head of Legal for the IFPI when they were pushing SOPA/PIPA, they had a real hard-on for these systems. And by 'just' I mean 'it closed on the 9th'
That said, I already included a bunch of examples of similar stupidity, including my own experiences with the white/pink noise claims. forgot about this one though, so I didn't include it.
[ link to this | view in chronology ]
FTFY
Purrr-fect.
[ link to this | view in chronology ]
Music company's make billions from YouTube they are not going to go along with shutting down content ID, what needs to happen is there needs to be some punishment for people who abuse the system to get money off videos they did not create or make dmca notices on obviously non infringing content . Content ID also protects YouTube from being sued by media company's it's not perfect it has obvious flaws. And wait til filters go up in the EU looking to scan all audio video and images that ll be a disaster for ordinary users who want to post memes or just a short funny video
[ link to this | view in chronology ]
Re:
And just how much of that goes to the artists whose works they claim are being infringed on. An external audit would likely make interesting and incendary reading.
[ link to this | view in chronology ]
Re: Re:
That's a different problem, though. Under the current copyright regime, if the artists signed over ownership of their copyrights under their label contract, then the label are the copyright owner that YouTube have to obey if they issue a DMCA takedown notice.
I doubt that any external audit would reveal anything that wasn't already well known during the 90s, when major artists like Prince and George Michael launched very public protests against how little money they were getting. The only difference between then and now is that there's more options for artists to go independent (or if they wrote their own music, re-record and release independently as numerous artists have done), but the basic "sign with a major label and you'll get an advance in return for your future artistic and financial soul" paradigm hasn't changed much.
I'm all for artists getting paid as much as possible, but if the reason they don't get much is because they agreed to sell everything for bottom dollar when they were starting out, there's not much the rest of us can do.
[ link to this | view in chronology ]
Techdirt: “It’s their site; they can moderate any way they like. If you don’t like it, create your own website. They’re not attacking free speech.”
Techdirt: “How dare websites pull down content on their own platforms? Content moderation is broken. They’re attacking free speech!”
Cognitive dissonance is real here, which is why I rarely look at this gaslighting website anymore.
[ link to this | view in chronology ]
Re:
The problem is that they are not allowed to moderate it any way they like, they are compelled to act on bogus claims under a framework in which the claimants have no liability for filing false ones.
[ link to this | view in chronology ]
Re:
Another day, another person inventing a fictional version of what's been said in order to pretend there's a problem.
YouTube aren't moderating the site as they see fit, they are doing what they're being ordered to do under the DMCA and problems arise when the people sending those orders are lying. If YouTube were allowed to make a moderation decision without being held liable for not doing so, the content would not have been taken down. This is perfectly consistent with other positions stated here.
"which is why I rarely look at this gaslighting website anymore"
Please continue not reading it, and go whine elsewhere more suitable to your limited reading ability, since the content here is obviously far too nuanced for your mind to process.
[ link to this | view in chronology ]
Re: Re:
DMCA does not require upload filters. ContentID is a voluntary development. Or it was until it inspired other laws (like that EU thing) that do require such dysfunctional censorship.
[ link to this | view in chronology ]
Re: Re: Re:
What the fuck are you talking about? Do you really think ContentID exists because of DMCA? YT was sued by Mediaset, Viacom and Premier League because they felt YT didn't do enough to prevent the uploading of their copyrighted material. The lawsuit was settled in 2014 with YT paying an undisclosed sum of money. ContentID is the logical extension of that lawsuit, it's way to avoid getting sued again and appease the rightsholders.
From the perspective of someone who doesn't understand context and the forces behind why companies do things it can be called "voluntary development" but the reality you are ignoring is that they were forced to do this or keep paying out billions in damages again and again.
[ link to this | view in chronology ]
Re: Re: Re: Re:
You'd have a point, if it wasn't that ContentID was developed in 2008
[ link to this | view in chronology ]
Re: Re: Re: Re: Re:
The lawsuit was filed in 2007 and settled in 2014. Before 2007 the media companies had been making noises about YT and copyright for a long time and how unhappy they where.
Do you understand now?
[ link to this | view in chronology ]
Re: Re: Re: Re: Re:
He doesn't have a point because the system created to mitigate ongoing lawsuits was done before the lawsuits were settled, even though part of the reason the settlement was agreed was because ContentID existed?
[ link to this | view in chronology ]
Re: Re: Re:
Yes it's a voluntary development to try and get from under the weight of the massive number of DMCA related lawsuits, including from people like Viacom who were suing them over content they uploaded themselves.
You have to be a special kind of stupid to think that this means they're not doing it out of necessity, or that obeying these legal orders means that defence of voluntary moderation elsewhere is hypocritical. But, here you are...
[ link to this | view in chronology ]
Re:
Please show consideration to your fellow commentors and clean up any excess straw after you're done whaling away at the strawman you've created.
[ link to this | view in chronology ]
Re:
[citation needed]
[ link to this | view in chronology ]
Re:
Rehab can help you with those hallucinations.
[ link to this | view in chronology ]
Re:
The assholishness is strong with you, feeble-minded one. Glad we shall be when visit anymore you don't.
[ link to this | view in chronology ]
Re: Womp Womp
Fuck your feelings!
[ link to this | view in chronology ]
Re:
"Cognitive dissonance...gaslighting..."
Let's add those two to the prodigious list of words you don't understand the meaning of.
Oh, prodigious means lots and lots.
[ link to this | view in chronology ]
Re: Re:
I once ran across a person that thought gaslighting meant you set your farts on fire...
[ link to this | view in chronology ]
Couldn't agree more.
[ link to this | view in chronology ]
Dude, it's obvious.
Clearly it was a copycat.
[ link to this | view in chronology ]
Music labels are simply parasites
Manipulating an algorithm that simply doesn't work.
[ link to this | view in chronology ]
Animal Creatiity?
If a monkey can take a picture and it becomes public domain, does public domain also own the cat purr? ("Arrangement by Fluffy, lyrics by Fluffy")
[ link to this | view in chronology ]
Re: Animal Creatiity?
If a cat took the video, yes. If a human took the video, then the human holds the copyright (if it's copyrightable).
[ link to this | view in chronology ]
Libel?
If I post a video about vehicular maintenance, upon which a robot's dart-of-persecution happens to land, and I suffer an income loss as a result, how is it that whoever runs these robots that continuously point their digital fingers and lob accusations of felonies aren't open to libel lawsuits? In what other circumstances is it perfectly okay to falsely accuse someone of federal crimes for the purposes of taking away their income? Restraint of trade? Anything?
[ link to this | view in chronology ]
Re: Libel?
"If I post a video about vehicular maintenance,"
Then, it should be silly if that gets taken down, but the first thing to look at is why. Did you use background music you didn't write or licence? Is the takedown notice coming from the music, or is it coming from the car manufacturer? Is it a complaint from another similar channel who are accusing you of something? The subject of your channel does not mean it can't infringe.
"how is it that whoever runs these robots that continuously point their digital fingers and lob accusations of felonies aren't open to libel lawsuits?"
Because although they seem random to the end user, they're not. They're making a best guess judgement based on data given to them by the major record labels and other corporate entitles for whom the modern copyright regime is designed. Whenever there's doubt, they will side with the people who sent them the notice to take down on the first pass.
"In what other circumstances is it perfectly okay to falsely accuse someone of federal crimes"
They're not accusing you of federal crimes, they're accusing you of a civil copyright infringement, and they're doing so based on the legal notice given to them by the original holder of the copyright you're accused of infringing (that is, it's actually the record label/whoever accusing you).
There are of course ways to respond to such a claim, but you should be aware that most of those thing (such as fair use) are defined as a defence in court, not a reason why YouTube shouldn't react to takedown notices.
"the purposes of taking away their income? Restraint of trade?"
They can't do any such thing. All they can do is withhold the part of ad income that they have agreed to pay you based on the popularity of the videos you host on their property. They can't affect the same video hosted elsewhere, and having videos taken down won't really affect your offline business unless you're intend on publicising it.
[ link to this | view in chronology ]
Re: Re: Libel?
You're needlessly complicating the question that is being asked: should bots that can't understand context and are even bad at identifying content have the power to decide if something is infringing?
Since you don't seem to want to answer it, you're deflecting by adding complexity that don't apply to either the case in the article nor the hypothetical from Rich.
They're basically random in results even if the algorithm has its logic. If you can identify a cat purring as infringing (other examples include white noise, bird songs and more), your algorithm is wrong and should not be used as a basis for accusations that have pretty dire consequences. Also, siding with the accuser is the very reason this whole process is unfair. In any legal process, doubt benefits the defender, for good reasons. And these accusers in particular have long demonstrated that they are not trustworthy, which makes it doubly unfair.
Let's put that back in the context of this article: no background music. Just the content as described. "Why" in this case is obvious: false positive in looking for infringement. Let's stop pretending that we should contextualize more than needed.
For some people, this is their business. (They might have another job, or another side to the same job, but maybe this is also necessary for them to reach a living wage... Who are you to tell that their ad income is not important?) And just three wrong notifications like this and the whole business goes under. I agree that making your business dependent on shaky foundations like a platform that can write you off arbitrarily in a few minutes is not a great business model, but it is theirs. Adding a layer of unfair non-judicial process on top of already unfair laws shouldn't be tolerated.
Great. Problem is the dominant position of Google on this market. No other platform will give the same degree of exposure. For those who want to make a business out of their videos without a heavy investment in time (which they might not have) and possibly money (which they might also not have), YT is the most obvious option. It's not perfect by any standard, but it's the best one for the general public at the moment.
[ link to this | view in chronology ]
Re: Re: Re: Libel?
(Note: I do agree with the point about "federal laws" though. That was a little off subject.)
[ link to this | view in chronology ]
Re: Re: Re: Libel?
"should bots that can't understand context and are even bad at identifying content have the power to decide if something is infringing?"
Typically, they don't because the bots are basing their decision on information given to them by other people. If the bot says "this video contains 30 seconds of song X therefore it's infringing", unless it made a mistake it's only basing that judgement on the people who told them that song X is under copyright and to be enforced.
But, the question is - who else should make the decision? Humans can make mistakes as well, and the whole point of the system is to avoid lawsuits from the people who issued DMCA takedowns over the content being used.
"Let's put that back in the context of this article: no background music. Just the content as described. "Why" in this case is obvious: false positive in looking for infringement"
OK, so a mistake happened. Mistakes can be rectified, and even though the process can be a pain in the ass I know of several channels who have had their movie reviews that extensively utilise clips from movies restored along with their income. You'd obviously prefer it if you didn't have to challenge the request in the first place, but mistakes happen and the system is in place because of people suing YouTube for copyright infringement they didn't commit.
"For some people, this is their business."
Then, they have gone into this business solely dependant on YouTube paying them revenue even though the system you're complaining about has been in place for close to 15 years at this point.
"Problem is the dominant position of Google on this market."
...which will remain indefinitely so long as people whine and whine about YouTube instead of using competing services. Catch 22. Those services exist, but nobody's going to use them as much as YouTube if all the content they want only exists on YouTube. Why would anyone go anywhere else if they can't get your content anywhere else?
Nothing will change if your reaction to problems is to put up with them regardless. You have to take action to change things. I subscribe to numerous channels that put up content on Twitch, Vimeo and even their own hosting platform, and set up their businesses to be less dependant on YT ad revenue and use other sources of income like merch, Patreon, etc. If they lose ad revenue due to a false claim, their business suffers but is not dead.
If you're too lazy to do the same, you're part of the problem if you suddenly lose your only source of income, even if the reason for that loss it a false positive. Same with any other business - if you run a bar and refuse to get supplies from anywhere but the one supplier, it's your fault if you run out of drinks if your main supplier screws up a delivery. Not your fault the order screwed up, but your fault that all your income is coming from a single source.
[ link to this | view in chronology ]
Re: Re: Re: Re: Libel?
Also worth noting that a business model dependant on the whim of a third party hosting service does seem a bit precarious from this point of view, but many people honestly don't believe that since they don't post copyright infringing material, they won't have any problems. You know, the ever popular "if you're not doing anything wrong, you're going to be fine" approach. As whacky as it seems, there are still some people out there that actually naive enough to believe that.
(side note: The book 1984 should never be banned or trigger warned in any school, like that dopey British university. 1984 should be mandatory reading for every student. Not "mandatory, or you don't get an A", I mean "Mandatory" as demonstrated in a Clockwork Orange, but I digress ).
The terms of service say that videos might get pulled if you post illegal things, and you might lose your earnings if you post illegal things, and they are also pretty clear that you could lose your account if you post illegal things, but nowhere does it say specifically and clearly that Google can yank the financial rug out from under you, and toss you off the platform just because there was a loud cat, white noise, or a clip from a stream you actually do own because some anonymous robot belched a completely unjustifiable objection to.
For what? I understand that illegal copies of movies or music can take a bite out of the wallets of the copyright holders, and that their are laws against it, and appropriate penalties can be levied against those who use copyrighted material without permission. What I don't understand is why there can be no penalties those who knowingly allow robots to do what they do and take money and possible audiences from perfectly innocent people actually respecting the law. (Before anybody pipes up about it, I do recognize that nobody actually saw that their little robotic dunce decide that purring was an infringement, and promptly gave an approving and encouraging nod, but they ABSOLUTELY do know that in their relentless quest to prevent anybody from hearing or seeing something without permission, there have been plenty of innocent casualties. The fact that whoever has the controls for these damn things has done nothing to reduce the absolute unquestioned faith they continue to promote in these things all but proves that they believe that the false positives and any collateral damage resulting there from, are acceptable losses given whatever benefits their might be. "Acceptable losses" means that you account and take some responibility for. You must accept some responsibility... it's right there in the name...acceptable losses. U.P.S., FedEx, etc., rack up millions in parking fines every year in New York city alone. Then, on an anual basis, they negotiate a lump sum payment, and that is an acceptable loss.
They don't just park whever the fuck they want and throw the tickets away while shouting "acceptable losses", because they actually understand that there is no reasonable way to run a large delivery company without getting tickets, so they plan ahead...."Jesus, will this ass never shut the hell up?"...good question, and "yes").
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Libel?
"As whacky as it seems, there are still some people out there that actually naive enough to believe that."
I understand why some people would believe that. I'm also aware of a lot of the problems that lead to "false" positives and why that standard doesn't apply. On the enforcement side, you have to remember that a lot of these issues aren't down to a bug in YouTube's algorithm - someone has actually sent a DMCA claim against something that is in the video, even if it seems ridiculous that it would be a valid claim.
"nowhere does it say specifically and clearly that Google can yank the financial rug out from under you"
No, it says so in the same terms of service. When you sign up you agree that your content will be scanned by ContentID, and that if that scan results in a positive for infringement then the same terms of service state that you won't get paid for infringing content.
The only way that Google can yank everything from you is if you're dependant on their ad revenue, nothing else outside of that is directly affected. We're strictly talking here about an agreement that person has made with YouTube and the share of ad revenue they pay out.
"What I don't understand is why there can be no penalties those who knowingly allow robots to do what they do and take money and possible audiences from perfectly innocent people actually respecting the law"
Because it would be insanely stupid if there were. First, the amount of content being uploaded to YouTube and the number of DMCA notices they get every day make it impossible to do without automation. Especially with the latter - if they get a new DMCA notice, they have to rescan their entire catalogues for infringing material. I'm all ears as to how you suggest that without automation, but fining people for using it is hardly going to help anyone.
As for the money - they agreed to YouTube's terms to receive the money which comes from YouTube. Why wouldn't they be able to refuse to pay people under such circumstances? I'll agree that there should be better ways of dealing with false positives than the hoops people sometimes jump through now and that they should restore any money potentially lost if they are found to not actually have been infringing. But, there's nothing untoward in them not paying people who have broken the contract that must be honoured to be eligible to get it in the first place.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Libel?
First...
"No, it says so in the same terms of service. When you sign up you agree that your content will be scanned by ContentID, and that if that scan results in a positive for infringement then the same terms of service state that you won't get paid for infringing content."
Sounds perfectly reasonable. Most people would read it that way. What might be more accurate "and if that scan results in a positive match, regardless of any actual infringing content, and with little to no explanation of anything...etc."
Second, I am completely lost about whether or not you are agreeing anything I said, or are arguing against anything I said.
For example, you:
"On the enforcement side, you have to remember that a lot of these issues aren't down to a bug in YouTube's algorithm - someone has actually sent a DMCA claim against something that is in the video, even if it seems ridiculous that it would be a valid claim."
I absolutely agree that if YouTube receives a valid takedown request, they are 100% obliged to comply. The meat of my original comment was not about accuracy of any YouTube algorithm, I was trying to understand why, if a person can be sued for libel if they make false accusations against someone that results in a loss of income or damage to reputation, are there no legal checks to keep some of these 3rd party algorithmically challenged robots from firing off false DMCA takedowns, in such a way that, as in my previous hypothetical, a person might be deprived of income, or have the reputation damaged because a video that included the hum and pur of a well tuned engine tripped some 3rd party algorithm which quite happily sent, in the form of a DMCA take down, an accusation of the commission of federal crimes.
To spell it out again, I understand that you can't hold an automated algorithm responsible for civil liabilities, but I don't understand why whoever is in charge of these 3rd party bots can't be made to answer for them when they do harm. I don't know much about law, but it would appear that if you ever want to actually publicly accuse someone of a federal crime to hurt their business or reputation, just write a bot to do it, that way, rather than you being taken to court, everyone would just throw up their arms, shrug, and say "algorithms are hard", and for some reason, everybody would just accept it.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Libel?
ContentID isn't accusing anyone of committing a crime.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Libel?
You are missing one very obvious thing, an automatic algorithm with the sole purpose of finding infringing content does not in any way cast shade on the person who uploaded/created the content regardless if the match was in error or not.
An algorithm specifically created to go after people in an attempt to discredit them means that the sole purpose of it is malicious. Ie what matters is what the intent was when the algorithm was created.
TL;DR: It's the intent that decides if something can be considered defamation or libel. ContentID's sole intent is to find infringing content, not ruin reputations.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Libel?
""I absolutely agree that if YouTube receives a valid takedown request, they are 100% obliged to comply. The meat of my original comment was not about accuracy of any YouTube algorithm"
You have to remember the sheer amount of data that's involved here. They already store more content than a person could ever hope to watch in their lifetime, and they're adding days of content every minute, then everything they already have needs to be scanned for new takedown notices. Even if the algorithm they use is 99.99999999% accurate, that likely still leaves hundreds of false positives every day.
That's not good and they could be a lot better at dealing with complaints and resolving conflicts in ways that don't negatively affect the accused. But, unless you have a better idea this is as good as it can get, especially if - and you have to remember this part - there's a history of the people telling YouTube what to remove lying to them. It doesn't matter how good your algorithm is if the data it's processing is wrong.
"in the form of a DMCA take down, an accusation of the commission of federal crimes.
Again, this is wrong. The infringement being accused is civil, not criminal.
I understand your points, but you seem to be getting angry over a rather misleading version of what's going on.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Libel?
"I understand your points, but you seem to be getting angry over a rather misleading version of what's going on"
A more accurate assessment would be "you seem to be annoyed that nobody is buying in to your deliberate misinterpretation of a hypothetical situation for the purposes of exploring a possibly rather asinine means by which the hypothetical victim of an automated (or intentional for censorship purposes, such has occurred on occasion and has been documented on this site in the past) DMCA requests, could actually put together a legal means to recoup losses. I don't think at all that every takedown request should be treated as an accusation of federal crimes, but as previous articles have pointed out, these automated systems have a much, much larger false positive ratio than anyone wants to admit, and I wanted to see what possible push back people who are getting the short end of the automated stick might have.
Personally, I agree with the W.O.P.R., aka Joshua: the only winning move is not to play.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re: Libel?
" but as previous articles have pointed out, these automated systems have a much, much larger false positive ratio than anyone wants to admit"
Citation? Maybe with references to how they're higher than the other solutions that are possible with the volume of data being processed. Or, are we arguing the toss because YouTube is only achueving 99.9999% accuracy rather then the preferred 99.999999%?
From where I'm sitting, there is a problem with automation. But, it's a problem with the automation of what's meant to be a legal claim under penalty of perjury, not the automation of YouTube dealing with the volume of legal demands it receives.
"Personally, I agree with the W.O.P.R., aka Joshua: the only winning move is not to play."
Then don't.
[ link to this | view in chronology ]
There's a copycat joke in here somewhere...
[ link to this | view in chronology ]
There's a copycat joke in here somewhere...
[ link to this | view in chronology ]
Re:
“I have the copyright on copycat [no citation needed]. Your money is mine now.”
[ link to this | view in chronology ]
Said before.
It would be nice to have 1 location that has ll aht CR and IP data to cross reference.
Esp with OWNERS NAMES.
Which would be cool also to tax revenue for those corps.
[ link to this | view in chronology ]
According to YouTube's Content-ID system both EMI Publishing and PRS own the rights to a 12 second purring loop.
So you know what? Let's you and them fight, and leave the cat out of it. Come see us when you're done.
[ link to this | view in chronology ]
Content ID claimed that a 1-minute video of insect noises I uploaded was copyrighted.
[ link to this | view in chronology ]
Re:
Probably jealous of all the buzz they were making.
[ link to this | view in chronology ]
Nuke the city if it will get rid of the terrorists in it? Sure.
The US Copyright Office foolishly considers automated copyright enforcement systems like Content ID as "standard" technical measures as defined under 17 USC 512(i). Automated copyright filters are not and should not be standard. They produce false positives (and false negatives) in abundance, and small online service providers wouldn't be able to implement them if the Copyright Office were to mandate them — which it may well seek to do in the near future.
Big publishing companies are willfully and maliciously blind to the fatal flaws of automated copyright filters. I've been hesitant to go as far as to say the same about the Copyright Office, but the Copyright Office has always skewed toward the interests of copyright maximalists at the public's expense.
[ link to this | view in chronology ]