Thursday, April 30, 2009

Removal of Lessig video apparently NOT a DMCA takedown

It was inevitable that reports of a "DMCA takedown" allegedly issued by Warner Music Group on a video presentation by Larry Lessig would garner lots of attention. I've refrained from commenting so far, because I strongly suspected that initial reports of a "DMCA takedown" were inaccurate, and I wanted to wait until some true facts emerged. Well, they have, and, as I suspected, this is a lot less dramatic than it seemed at first glance. In sum, it appears that initial reports were wrong; WMG apparently did not issue a DMCA takedown notice on the Lessig video.

Professor Lessig explains on his blog what happened:
As you may have read me tweet, the organization that hosted me for this talk... [r]eceived a notice that Warner Music had objected to its being posted on copyright grounds. Apparently, YouTube's content-ID algorithm had found music in the video that they claimed ownership to. The organization is apparently responding by disputing the claim. I'll report back when I hear more.
The key is Lessig's statement that "YouTube's content-ID algorithm had found music in the video." Here's what that means: YouTube employs an automated system for identifying music uploaded to its system. The major music companies, in cooperation with YouTube, have uploaded digital "fingerprints" of their songs to YouTube's content ID system (which uses AudibleMagic's technology). When a user uploads a video, the content ID system checks the music against the database of fingerprints, and applies a "business rule" set by the copyright owner: 1) block, 2) track, or 3) allow, and share revenue. Most major record labels participate and use option 3. WMG did so until late last year, when its agreement with YouTube broke down, and the setting for WMG's music flipped from "allow and share revenue" to "block."

So today, when a user posts WMG music to YouTube, its content ID system identifies it, blocks it from being displayed, and sends a notice to the user. At this point, the user can dispute the block through a simple online form. This is not the DMCA notice-and-takedown (and counternotice) process. Rather, it is a private system -- totally outside the DMCA -- that YouTube established to allow it to monetize videos (and mollify copyright owners, who are relieved of some of the burden of policing YouTube for infringing copies of their works).

So, assuming that Lessig accurately described what happened here (and I have no reason to doubt him), the video of his presentation was not the subject of a DMCA notice. (Indeed, Lessig's initial Twitter report did not claim this was a "DMCA" takedown.) Rather, the video was identified by YouTube's filter and blocked, and the copyright issue can probably be resolved by the poster (an organization that Lessig does not identify) taking 5 minutes to fill out an online dispute form. If, at that point, WMG still believes that the use of its song was infringing, it can then issue a true DMCA takedown notice (and, if necessary, the poster can file a DMCA counternotice).

As to the underlying issue: was Lessig's use of WMG's music a fair use? The problem I have in answering that question is that I'm not even sure which music triggered the video's removal. Lessig lists the music incorporated into his presentation, but the list makes no mention of WMG. (Was there simply a malfunction in the content ID system?) However, having watched the video, it does seem that Lessig used only brief excerpts of music to illustrate his arguments about "remix culture." In other words, probably fair uses (though, again, I'd want to know for sure which song was the subject of the claim before saying so for sure).

So what will happen? If the poster disputes the takedown through YouTube's system (which Lessig indicates is occurring), my guess is that WMG will not press the case, and the video will get re-posted. YouTube's content ID system isn't perfect, but it does have a feature that permits disputes like this to be resolved quickly, and without much hassle -- if that's what the parties actually want.

UPDATE: Lessig reports on his blog that "the protest filed by the uploader to the block was successful." Should be the end of the story.


  1. Finally some clarity about what was a confusingly reported story. Thanks for following up on this. I think a lot of people got the wrong idea about whose actions were the impetus behind this "takedown". Looks like it actually was Google.

  2. This is not the first time a prominent person has been hit by YouTube’s fingerprint identification system. James Randi (aka “The Amazing Randi) had his account suspended for several days for the very same reason.
    Here is dprjone’s video announcing Randi’s suspension:

    And here is Randi’s explanation as to what happened:

    Notice that Randi’s account was restored in 3 or 4 days because the copyright owner was a friend of the JamesRandi foundation and he asked YouTube to restore Randi’s account. One wonders how long it would have taken to get the account restored had this occurred to an ordinary YouTuber? And what if the copyright owner didn’t like what Randi had to say? Why should he have the ability to keep Randi’s account suspended via YouTube’s fingerprinting system? If there was a legitimate issue over Randi’s use of his material, then he should HAVE to file a DMCA in order to get Randi’s account suspended. At least that way he would be legally and morally responsible for the suspension. And if Randi’s use was an obvious Fair Use, then the copyright owner could at least theoretically be held accountable under Section 512 (f) for filing a bogus DMCA.

    Why is account suspension necessary do to a digital match even necessary—especially in cases involving fair use? It simply isn’t. And YouTube’s lawyers know that this isn’t required by law, as pointed out by EFF’s Fred Von Lohmann in the following video (see 38:43 – 41:10)

    Here is what I THINK is happening. I THINK YouTube installed their digital fingerprint matching system so that they can go into court and claim to be cleaner than clean when it comes to honoring and protecting copyright. I THINK they want to this to get them off the hook in the Viacom case. I am not sure what good this does them. If the Court determines that their earlier behavior does not fall within the DMCA’s safe harbor, then they don’t need this system to cover their ass. And if the Court determines that their earlier behavior is NOT covered by the safe harbor, then using this digital fingerprinting to get them off the hook isn’t going to work. It is like getting caught driving 100 miles per hour and hoping to get off easy by changing your driving habits and driving only 30 miles per hour on the highway instead. This doesn’t diminish the original offense, but it does create a new hazard on the highway. If YouTube is found guilty of copyright infringement for their past behavior, then the filtering system will not diminish their guilt. It will, however, stomp on our right to make Fair Use of other people’s works.


Comments here are moderated. I appreciate substantive comments, whether or not they agree with what I've written. Stay on topic, and be civil. Comments that contain name-calling, personal attacks, or the like will be rejected. If you want to rant about how evil the RIAA and MPAA are, and how entertainment companies' employees and attorneys are bad people, there are plenty of other places for you to go.