It’s not unusual for anti-piracy companies to promote their takedown prowess based on their ability to send out millions of takedown notices every week. The problem, for some, is their deployment of autonomous robocops that give recipients 20 seconds to comply but can’t handle a complaint or muster a basic reply.
From: TF, for the latest news on copyright battles, piracy and more.
Faced with millions of instances of copyright infringement every day, many rightsholders use anti-piracy companies to help stem the tide.
More often than not, that involves sending DMCA takedown notices on an industrial scale, in the hope that Google and Bing delist infringing URLs from search results before the cycle begins again.
Huge volumes of DMCA notices and similar requests are handled directly by companies including YouTube, Twitter, Facebook and Instagram. And that’s just a part of a very large iceberg, much of it completely and permanently hidden, and almost all of it automated.
Trust the Machines
For years TorrentFreak has documented the most newsworthy takedown demands from the billions of notices sent to Google and other platforms with transparency programs.
Our focus is usually on the most egregious examples of wrongful and damaging takedowns, especially those that could’ve been avoided by taking basic steps guided by common sense. Programmed by humans yet blamed on machines, the robo-blunderers’ mistakes are rarely even acknowledged by those ultimately responsible.
Our own experience shows that URLs on this very website have been wrongfully reported to Google almost 150 times by 65 reporting organizations and rightsholders. We were targeted again in February for referring to a piracy app by name in an article reporting that its alleged creator had been arrested.
RoboCop: Book him!
Sgt. Reed: What’s the charge?
RoboCop: He writes about pirates
A full 7% of all wrongful DMCA notices filed against the torrentfreak.com domain are duplicate attempts to take down the same non-existent infringing content that caused the first set of notices to be rejected. “We have a good faith belief” means almost nothing the first time; after another futile attempt expecting a different outcome, it’s at best sarcastic, at worst a flat-out lie.
The Personal Touch Via Email
Rather than complaining directly to Google about infringements that don’t exist, sometimes anti-piracy companies communicate with us directly instead. More accurately, we receive emails from DMCA robocops (minor redactions, matter ongoing) that present as human but cannot complete rudimentary tasks.
Of course, there was no infringement, and in this case, that was made extremely clear by the sender’s own notice. It listed three URLs with domain names that are not torrentfreak.com and have nothing to do with us.
Despite that, DMCA robocop told us upfront that the legal document before us was accurate. “We hereby declare that the information in the notification is accurate to the best of our knowledge & belief,” the notice declared.
Now read that again in Robocop’s voice with “Stay Out of Trouble” tagged on the end.
I’m What You Call A Repeat Offender
This wasn’t a one-off; we’ve received these notices several times before. We’ve responded with emails explaining that their notices are wrong and also sent lengthy responses (complete with diagrams) explaining exactly why they are wrong and why we now need written confirmation that their legal complaints have been retracted.
No responses are ever received, but there is no question that resources were available for the same bots to keep sending out even more notices. “I Have To Go. Somewhere There Is A Crime Happening,” bots probably said as they clumped off into the distance.
Who Cares if it Worked or Not?
Obviously, our multiple cases are not isolated examples. Erroneous robo DMCA notices are reported daily, with some users having to report problems that originated on YouTube, to YouTube’s account on Twitter.
Why? Because the companies sending the notices simply walk away and refuse to listen to the people whose lives have just been turned upside down. In some cases, targets no longer have an account on YouTube as a direct result of the complaints yet are instructed to contact the claimant directly to resolve the dispute.
In our experience, mass senders of DMCA notices do not answer the people they target, whether they’re contacted on YouTube or directly by email. We know that because we sometimes try to get companies to respond to complaints by sending questions to the same designated email addresses.
Complaints Are Deliberately Impeded
We can’t go into real detail for legal reasons, but recent DMCA complaints took down an entire channel on Spotify yet the notice sent to the channel owners didn’t list an artist or an identifiable reporting group; basically, ‘robocop_enforcement’ was responsible, whoever they were. Worst still, the notice didn’t even explain whose copyrights were allegedly infringed or how.
Five days were given to contest the claim based on the information above, or those targeted were to be considered as consenting to the takedown. In the meantime, the flimsiest of DMCA-style takedowns was considered authoritative because someone’s DMCA robocop walked in, and when Spotify asked about the charge, it said: “Aiding and abetting a known felon.”
This is not “just a glitch” this how copyright law plays out on a daily basis. Nobody is accountable except for the machines and just to be clear, those machines are completely deaf, disinterested and unaccountable. The humans behind the machines are few, distant, and massively reluctant to have a sensible conversation about the damage they’re doing on a regular basis.
A personal theory is that since these companies deal with so much infringement, every individual, site or service they come into contact with is considered guilty by default. That is very convenient when an entire operation has been built around sending DMCA notices, but has no appropriate infrastructure to discuss the damage they’re causing.
I appear to have suffered an emotional shock. I will refer myself to a copyright crisis center
Image credits: Pixabay/SponciaHassan
From: TF, for the latest news on copyright battles, piracy and more.
Android 13 scheint sich schneller als die Vorgängerversion zu verbreiten. Vor allem seit Januar 2023 ist der Sprung deutlich. (Android 13, Smartphone) 
US-Geheimdiensten war die Existenz von zwei weiteren chinesischen Spionageballons bekannt, wie die über Discord geleakten Pentagon-Papiere zeigen sollen. (
Rovio ist für eines der erfolgreichsten Mobile-Games aller Zeiten verantwortlich - nun könnte Sega die Angry-Birds-Macher kaufen. (
Forscher haben mithilfe von maschinellem Lernen ein zuvor veröffentlichtes Bild eines schwarzen Lochs gestrafft. Zum Einsatz kam die KI Primo. (
KI-Sprachmodelle sind mehr als Wortkombinationsmaschinen. Sie können schlussfolgern - und auch denken? Auf jeden Fall werden sie unerwartete Fähigkeiten entwickeln, auch ohne dass sie ihnen einprogrammiert wurden. Von Thilo Hagendorff (
Wie viele Berufsgruppen befürchten auch Programmierer, dass eine KI bald ihre Arbeit macht. Statt sich zu sorgen, sollten sie sie lieber sinnvoll nutzen. Denn das geht. Von Christian Heilmann (
Ab 2.000 US-Dollar bieten Kriminelle an, legitim aussehende Schad-Apps in Googles Play Store zu schleusen. (
You must be logged in to post a comment.