The Problem With YouTube’s Content ID System…

For a while after YouTube started, the only way for a copyright owner to protest his material being used without his permission was with a formal DMCA takedown notice.  While a DMCA takedown notice is the most efficient third party method leaving the least room for abuse, it doesn’t really work unless you know that your work is being infringed upon.  This is why Google developed YouTube’s Content ID system.

Originally, the Content ID system was flawed, but better than it is now.  The way it worked was copyright holders would submit their work to YouTube’s system, and it would scan uploaded videos, as well as crawl the existing videos from time to time, for the work submitted.  Exactly what criteria it uses and how exactly all this works is kept secret by Google.  They say it’s to protect company secret and copyright holders’ work, but I fail to see how keeping this a secret accomplishes that end.

If it found something, what happened next depended on the copyright holder.  While a takedown notice was possible, more likely is that the copyright holder wanted to monetise your video for himself, either by including a link to purchase the material you’re using or putting advertisements on your video.  Now, because Content ID is prone to make mistakes (and the company secrets prevent us, the users, from pinpointing exactly why these mistakes are made), a user could dispute the Content ID match, and YouTube would rule in the user’s favour by default.  If the alleged copyright holder still wanted to claim the rights, they would have to file a formal DMCA takedown notice and take all the legal responsibility that came with it.  The uploader could also proactively file a DMCA counter-notice to force the claimant to put up or shut up.  But, that was then.  This is now.

Now, if a user disputes a Content ID match, a notice is sent to the copyright holder, informing them of your dispute.  Their Content ID match remains on your video until they release their claim or until 30 days have passed, whichever comes first.  I don’t know what happens if they don’t release their claim.  I’ve never had that happen.  I know what should happen.  They should have to file a formal DMCA takedown notice and take all the legal responsibility that comes with it.

So, I have a scammer called Myuap that’s been sitting on a dispute on my Content ID match for three weeks now.  They don’t plan to respond to my dispute.  They can’t legally claim the video as theirs, so they won’t.  If they release their claim, they’ll stop earning advertising revenue off the public domain.  So, they just plan to milk it for 30 days and move onto some other sap and repeat the process.

Google knows about scammers that abuse the flaws in the Content ID system.  If I were more cynical and not contractually forbidden from saying bad things about Google, I might suggest that the changes to the Content ID system were to accommodate Content ID fraud from “companies” like The Orchard Music.

So, how does this Content ID fraud work?  Well, the scammer submits a bunch of public domain pieces to the Content ID system and, when there’s a match, they steal from the public domain by collecting advertising revenue on someone else’s video.  On rare occasions, they’ll use the uploader’s own copyrighted material, stealing from the uploader.  They also steal from the uploader if the uploader planned to monetise the video in question and can’t because of a Content ID match.

I said before that Google knows about these scammers and refuses to get involved.  They refuse to enforce their own policy here, citing, “We don’t mediate Content ID disputes.”  That’s a red herring, of course.  If I were more cynical and if I weren’t contractually forbidden to say bad things about Google, I might suggest that Google keeps them around because they make plenty of money enabling these scammers’ theft from the public domain.

Obviously, the Content ID system is a good concept.   As a copyright holder, I’d like to know about infringements on my work as soon as I possibly can. It’s just horribly implemented.  I have some ideas to improve it.

  1. Take the Content ID system down for a couple weeks to purge all illegal material from it. Send in spiders to find public domain work, purge all public domain work from the Content ID system, and ban all users who submitted public domain work (with the exception of those who owned it when they submitted it and subsequently released it, unless it’s a recurring problem with those people). If they’re going to use the Content ID system to defraud, then they’ll just have to file formal DMCA takedown notices if they have a legitimate copyright claim.
  2. Shift the burden of proof to the accuser.  If the Content ID system matches content, and the uploader disputes it, immediately restore it to its previous condition, but send a notification to the alleged copyright holder that includes a link to the matched video and the entire context of the dispute. That way, if there is a legitimate match, the copyright owner is well equipped, but we’re not convicting without evidence.
  3. The dispute process should have a way of marking whether the uploader believes this match is fraud.  Any disputes marked as fraud will be investigated by Google staff, especially several disputes against the same claimant.  If fraud is determined to be the case, the claimant may lose the privilege to use the Content ID system.  If a particular uploader has been known to knowingly falsely claim fraud on too many occasions for it to be an accident or mistake, he might also be subject to losing his account.  This feature is to maintain the purge of public domain material, as well as to purge copyrighted material attributed to a person or organisation other than the legitimate copyright holder.
  4. Allow the uploader to file a formal DMCA counter-notice and take all legal responsibilities that come with it at any point after the Content ID match was initiated.  This ballsy move says, “I know I’m right and I’m prepared to go to court over it.  If you are, then file your formal DMCA takedown notice now.  If you aren’t, then drop it.”  In clear cases of abuse of the system or fair use, this will expedite the process.  In cases where the uploader is wrong, then he wouldn’t want to do this and, if he does anyway, he brought the problems on himself.
  5. Give the alleged copyright owner more options.  As a copyright holder myself, these are the options I (and dare I say most copyright holders) would like to be given:  First of all, the option to drop it is probably going to be the most used, which would require no further action on anybody’s part.  It’s the default position.  Second, I would really like the option to negotiate with the uploader.  In my experience, negotiating with your infringer is far more effective than getting a legal system involved.  If they select this option, then Google opens up a mediated form for both parties to communicate.  The only purpose for Google’s mediation is to confirm the agreement of this negotiation so lying is impossible, though I do say they should be able to get involved if they want to, but should be under no obligation.  If this method proves ineffective, then the claimant could move onto one of the other two options.  The third option is filing a formal DMCA takedown notice and taking all legal responsibilities that come with it.  Reinstating the effects brought on by  the Content ID match will not be an option unless agreed upon in negotiation.
  6. Once a formal DMCA takedown notice is filed, I have no issues with how Google handles it from there, but then I only know the policy.  I don’t know if that’s the way it works in practice.  I’ve never had to be on either end of such a notice.  I’ve always gotten the infringer to take down the offending material by a simple request.

I believe Google already knows everything I said in this article.  I just don’t think they care.  I’ll send them a link, just in case they care, but don’t hold your breath that they’ll be interested in making YouTube a better place.

There are 2 Comments to "The Problem With YouTube’s Content ID System…"

  • silver accountNo Gravatar says:

    The new appeals process and improvements to content matching are welcome additions to YouTube’s Content ID system, but they may have potentially removed the bite out of the manual review process. Leaving it up to the content owners could prove troublesome. Of course, the content owners will still have to file a DMCA takedown notice if the video creator appeals the decision after the manual reviews process. This puts more power into the hands of the creator to prove that their video doesn’t infringe any copyright.

  • Joline SilveriaNo Gravatar says:

    Can I just say what a reduction to seek out someone who really knows what theyre speaking about on the internet. You positively know how you can convey a difficulty to gentle and make it important. Extra individuals need to learn this and understand this aspect of the story. I cant imagine youre not more in style since you positively have the gift.

Write a Comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Shortcuts & Links


Latest Posts