Google's daily Transparency Report data-dump includes all DMCA requests

Fred von Lohmann, Legal Director at Google, has published a blog-post explaining the company's new practice of publishing data and reports on the number of takedown requests they get. It's all about helping policy makers understand whether the censorship provisions in the Digital Millennium Copyright Act are doing their job:

Starting today, anyone interested in studying the data can download all the data shown for copyright removals in the Transparency Report. The data will be updated every day.

We are also providing information about how often we remove search results that link to allegedly infringing material. Specifically, we are disclosing how many URLs we removed for each request and specified website, the overall removal rate for each request and the specific URLs we did not act on. Between December 2011 and November 2012, we removed 97.5% of all URLs specified in copyright removal requests.

As policymakers evaluate how effective copyright laws are, they need to consider the collateral impact copyright regulation has on the flow of information online. When we launched the copyright removals feature, we received more than 250,000 requests per week. That number has increased tenfold in just six months to more than 2.5 million requests per week today. While we’re now receiving and processing more requests more quickly than ever (on average, within approximately six hours), we still do our best to catch errors or abuse so we don’t mistakenly disable access to non-infringing material.

More data about copyright removals in Transparency Report (via Copyfight)


  1. Two frekin million a week?  How big would the cube-farm have to be to manually process them all?  How many workers, generating no revenue for Google? The only good news here is that infringing videos are being posted ten or a hundred times faster.

    1. That’s 300,000/day, which is about 3/sec, which is not all that much, technically. A laptop could process that many.

      EDIT: derp, missed the word “manually”

      1. It’s not a terribly significant tech problem (especially for an entity like Google, that has a certain amount of in-house expertise in cheap data-shoving); but it sure helps demonstrate why even the shoddiest and most implausible DMCA takedowns almost always work:

        Even cursory human intervention would be wildly, unbelievably, expensive(and there isn’t much incentive to provide it; but it’d be a serious logistical challenge even if there were).

        Consider, for context’s sake, that the total number of filings in all US District Courts, as reported by the judiciary in 2011 was 367,692.

        If we were attempt to provide DMCA takedowns with something equivalent to judicial process, Google alone would be turning over a year’s worth of federal caseload every 30-ish hours.

        As a database ingest issue, it’s trivial; but it demonstrates that we’ve managed to build an IP system sufficiently gigantic that there probably aren’t enough lawyers on the planet to operate it on any basis other than automatic assumption of guilt…

        1. This. This right here. Shared article with your comment as the enticement text. Attributed and all but… dude I can’t think of anything meaningful to add to the convrosation.

  2. Just a question here. Since there is no penalty for filing a bogus DMCA request, why isn’t some activist group filing DMCA takedown requests on the entertainment industry? What would happen if some group started sending in a couple of dozen takedown requests on every new movie promotional website that went up? And file another couple of dozen on the same site when it came back up?

    1. It’d be worth doing for the lulz (and it might even work against the less visible promotional arms of major media entities: social network astroturf that includes copyrighted promo material, random ad agencies who are in fact working under contract but aren’t trivially verifiable, etc.), however, I suspect that you’d quickly find that DMCA takedowns are ‘automatic’ in rough proportion to how important a customer you are. 

      If I’m a web company who could lose my Safe Harbor protections for ignoring a DMCA request, and I get one against some free user who calls himself ‘fuzzyfuzzyfungus’, it is vanishingly unlikely to be worth it to run it by my lawyer, or get in contact with the user and see if he wants to contest the takedown before it happens, or similar. If, on the other hand, I handle a web hosting contract for Sony Pictures Entertainment Inc. it is almost certainly sensible for me to leave the DMCA takedown in ‘processing’ status while I confirm that the customer is willing to contest it, at which point I can dismiss the complaint with zero interruption. Similar things are likely the case for high-profile youtube accounts (after that Mars rover incident, say, I suspect that NASA’s account is specially blessed against robo-takedowns in a way that the little people aren’t).

      It would be an interesting scorched-earth campaign against astroturf marketing(much of which relies on both copyrighted material from the mothership and no visible ties with the mothership, which would make it a soft target), and might also be effective against weaker, smaller, companies, or contractors for large companies who aren’t well known to be such; but I’d be pessimistic about my chances against a big player.

  3. 97.5% ? I have this strange feeling that the only situation when Google doesn’t pull down the sites, is when randomly flailing copyright holders attempt to commit seppuku and target themselves.

  4. I think an obvious use would be to use these figures to find out what “experts” have the worst rates.
    The cartels always refuse to put much effort into protecting their “valuable” assets and have hired the lowest bidder for robotakedowns.
    Google is now having to devote actual resources to stay not only within the law, but verify these claims.
    Why is everyone else forced to carry the water for the cartels?
    Why is the law weighted to protect the cartels, even when they are out of their damn minds?
    If Google were able to bill the cartel $5 for each and every single bogus url in one of these takedowns, do you think they would improve the widely cast net?

    I think it is high time we embarrass the cartels.
    They claim its easy to stop piracy, as long as they don’t have to put much cash into it.  There needs to be a cost.
    They claim under penalty of perjury that the notices are believed to be valid.  The wording needs to change to require them to face the music when they keep doing a shitty job.
    We need to show the numbers of how much of this DMCA traffic is a waste of resources for the recipients, because the cartels don’t care that much.
    Oh and a special report on how many times they keep trying to get themselves delisted, because that is the ultimate statement of how little they actually care if they are following the law or not.

    1. The cartels have special access to submit.  They are afterall, the only “people” allowed to hold copyright.
      Heck UMG even has their own special take-down button… see UMG vs MEGA.
      The question isn’t what other cartel members have that special button, it is why it exists in the first place.

Comments are closed.