More cracks in YouTube's takedown process reveal how media giants and corporations get to claim copyright to things they don't own

An unsigned rap group called After the Smoke couldn't post their song "One in a Million" to YouTube because every time they tried, it generated a YouTube content-match error saying that Universal Music owned their song. It turned out that UMG had laid claim to a leaked video that had a UMG artist performing the unsigned band's track in it, and this effectively gave Universal the power to censor the unsigned band's song.

YouTube's content-matching system has a lot of problems, as archivist Carl Malamud discovered when corporations started to claim that they owned the public domain US government videos he posted, threatening to cost him his YouTube account. And Universal attained notoriety for abusing content match by claiming to own the song that MegaUpload commissioned from major artists criticizing Universal and other rightsholder groups for their copyright stance.

Universal Music May Have Inadvertently Exposed a Flaw in the YouTube Takedown Process


  1. It is not so much a flaw in the content match system but in copyright itself.
    Google is left in a position having to assume that only corporations can hold copyright, otherwise they face huge lawsuits and protracted battles wasting time and resources.  This gives large copyright holders huge advantages to shut out content they just don’t like, with no downside to them.
    The law is written by and only benefits the corporations that hold copyright, they can take down anything claiming they own it causing harm to the actual owner and the punishment for doing that?  Nothing.

    1. It is a flaw in the system, because it includes only a small handfull of people who can register their content. If we could all register our content, it’s be some kind of fair. Also it’d be dead pretty soon because nobody at all would like it anymore, but that’s kind of the whole idea.

      1. The entire concept has been turned into something that only benedits corporations.  They extend the length of time things are covered and break the original deal, there was X amount of time and then the trade off was the monopoly on that was over as it became part of the public domain.  Now they have gotten the right to gut the public domain and take things back and lock them away again.  It was bad enough when Disney was ripping stories out of the public domain and making them their own suing anyone who wanted to use those same public domain stories, now they can wave a wand and just steal the culture to make a few more bucks and offer us nothing but more of the same.

        1. It was never really a deal to begin with. And it hasn’t been a deal eversince. Copyright, as a whole, all of it, is obsolete. We’re living in the age of ubiquitous cost free copying and distribution. In fact, it’s so embedded into our way of doing things now, that copying isn’t even avoidable, you are forced to do it everytime you press a key, enter any URL, consume any content etc.

          The entirety of copyright was invented in the age when copying was so difficult and expensive that regulating it seemed like a pretty practical idea. It’s not anymore. It’s now an insane idea.

      2. Without stricter and consistent penalties for abuse that would only expand the problem, as it would then simply empower everyone to assert claims to things they don’t own and enact sweeping censorship based solely on specious, unverified claims of ownership.

        The entire problem is that the law as it is written places the onus of evidence is upon the accused to provide an argument for why they shouldn’t have their content censored.

        No matter what content you would like to post online, I can have it removed without volunteering any evidence. Then, either you must accept the removal, or provide a written statement that I do not have a legitimate claim, and then even though my claim may be entirely fraudulent, you still must wait a minimum of 10 days to have your content restored.

        The law as it is written is very friendly, in fact quite suspiciously friendly, to enacting completely unjustified censorship of people by making openly fraudulent claims of copyright ownership.

        The system in use by Youtube is simply the logical extension of a legal framework which is itself custom tailored to be easily abused with impunity.

  2. I have to assume there’s something wrong with the match system too. Couple years back I made an ordinary video of nature photography using a recording of my daughter playing a Beethoven piano sonata. It got grabbed, I spent better part of a year writing to Google complaining until it finally went away and the vid played with the soundtrack. This thing it was:

    1. That’s because Beethoven registered all his music on youtube so you can’t upload videos containing his music.

      No, actually it’s not an error, it’s a perversion of the copyright system. See when Beethoven gets performed by an orchestra, they get to stamp their own copyright onto that recording. Then that orchestra sells their recording to EMI, and they register their recording on youtube. So from henceforth everybody who uploads a video featuring Beethoven will get taken down automagically.

      Wait, you’re right, it really is a fucking bug in the system. I mean, the system is the fucking bug.

  3. Youtube had that thing lately, how an hour of material is uploaded every second. I would love to hear how much material is taken down every second.

  4. It is a shame that there isn’t something like SOPA/PIPA to protect this sort of copyright infringement.  Wait, what?  SOPA/PIPA would have just empowered corporations to continue to flout creator’s rights even further?  Huh.  Weird, now I’m starting to think that SOPA/PIPA isn’t about protecting copyright at all!

  5. Yesterday my wife uploaded a video of our budgies taking a bath in a cup of water.  There’s some indiscernable muffled music playing and the only logo is on a bottle of Febreze in the background.  Her video was immediately flagged as a copyright violation.  It’s obviously a false positive, but what’s with this guilty-until-proven-innocent bullshit?

    Are we under arrest for this?

    1. The obvious solution is to stop consuming media and consumer goods and see how that works for the businesses that are making complaints. They are trying to get us to stop doing business with them, aren’t they?

  6. Let me tell you a little something about YouTube. They released this ‘YouTube for Education’ thing recently, and as part of our work developing internet filtering software, we have attempted to integrate the feature into our services as a way to provide ‘safe’ YouTube content to schools (safe as defined by YouTube, not by us).

    Their current implementation, as is usual with new software, is full of holes and flaws. We can see several easy ways to fix this, but good luck getting in touch with YouTube to discuss it with anyone. If you’re lucky, you can get their phone number, but you’ll never get past their front office to speak to a technician. They are a completely isolated organization who do not want to communicate with their users in a two-way fashion. And I understand why, since the volume of data they handle is simply impossible to handle otherwise. The only way they can keep their heads above water is to automate everything, and kowtow to the people who can destroy them (the content conglomerates) as much as is necessary to keep them off their backs.

    And it’s strange, because Google is much easier to work with. Run Google queries through squid, tack on ‘&safe_search=yes’ onto everything, and you’re done. But YouTube has come up with this crazy header system that you have to sign up for and is burdensome to maintain. Half the people we’ve beta-tested it with don’t feel it’s worth the effort.

    In a way, YouTube seems like the bastard child of the Google family, and it is clear to me all the big engineering brains in Google have not succeeded in getting YouTube to do things well. YouTube seem determined to do their own thing, no matter how sloppy or unfair to the end user. And perhaps, that’s a strategy. I would not be surprised if, when pressure got too great, Google would cut YouTube loose and let the RIAA/MPAA eat it alive, and start over with something more structured like Vimeo that mandated original content through a human-run arbitration process – which would also not be perfect, but at least you would have someone to reason with, instead of:

    Machine: You’re infringing!

    Me: No, I’m not!

    Machine: Yes, you are. End of line.

  7. Ah yes. The amazing YouTube content matching system.

    I uploaded an hour long video of me talking and honing a blade. There wasn’t even audible music in the background, just my voice and ‘scrape, scrape, scrape’. It was flagged as belonging to one of two entities (it changed, as I kept working with the video). It was ultimately unflagged without a ton of effort on my part, but I had to keep saying ‘NO, BMI does not own the ‘music’ in this video.’

  8. I’ve had quite a few videos get pop up with infringement notices – for songs I’ve written myself.

    When I try to argue my case, I am asked to fill out information like: 
    record label, band website, album title, year of release, etc.

    But I don’t have any of these things!
     I’m just a guy who picked up a guitar and a keyboard and put together something to provide background noise for my pointless video.  

    It is immensely frustrating how technological systems automatically assume guilt, and put pressure on the accused to prove their innocence.  
    It is fundamentally unfair, and goes against the established legal systems that the Western world has been using for hundreds of years.

  9. This is why we need a three strikes law. Three false take down notices and you can never send a take down notice again. Obviously, this needs to apply to the parent corporation, not the subsidiary.

  10. For all the amazing things that youTube provides, it fails in so many other ways. The comments are cesspools, nudity is out of the question, and it is rife with takedowns and stripdowns – with ALL the rights in the hands of deep-pocketed content hustlers and virtually no chance for small content producers (aka, you and me) to appeal their judge-jury-executioner style decisions.

    But I mean,… there are other video sites on the internet – are there some major reasons (such as functionality) that prevent users from ditching youtube in favor of more free alternatives?

    1. There’s an easy fix for the comments. Uploaders can delete them. I do all the time on the stuff I upload. And then I get a bunch of messages about how I’m infringing on their “free speech”. Which is hilarious since I’m on my third YouTube account due to users flagging my videos.

    2. nudity is out of the question…

      Not if you put one person in a lab coat and make it look like a doctor’s office.

  11. There is a correct way to handle this.  But YouTube doesn’t do it.  When you hit the end of the line with the robot match system, you should be able to send a DMCA counter-notice and have the content put back up, unless or until there is a court ruling.  YouTube even has an automated mechanism to receive counter-notices.  The problem is that they don’t honor their legal obligations in relation to counter-notices.

  12. We need to set up a video-sharing site hosted in a country that won’t play nice with US authorities (Russia?). Call it anytube[dot]ru and host anything that youtube repeatedly deletes. America’s politicians seem hell bent on breaking the internet for Americans… that doesn’t mean the rest of us have to put up with it.

    Anyone know of any reliable Ruski hosting options? I am serious about this.

Comments are closed.