In a Boing Boing article published earlier today, they write about Sony falsely laying claim to Bach's music on Facebook causing anyone who plays music from the composer who has been dead for over 300 years to receive a copyright dispute claim. The linked article incorrectly says this happened on YouTube, but it was Facebook.
an online legal system where you're guilty until proven innocent, coming soon... image from pixabay
James Rhodes a pianist Tweeted that a dispute was lodged via Sony automatically on a video he posted on Facebook, claiming they owned 47 seconds of his performance which was an original played in his living room.
This is alarming.
And it's not the first time an algorithm has incorrectly flagged content or a system so broken has allowed companies to monetise other peoples content by lodging a dispute, YouTube is perhaps the biggest serial offender when it comes to false content ID matches inaccurately flagging copyrighted content.
Famously in early 2018, a story was published about a professor who put up ten hours of white noise on YouTube and received five copyright requests on it, even though it was an original creation.
Why is this a big deal?
On September 12th, European Parliament is going to take a vote on a very controversial part of the new Copyright Directive where they'll decide whether to include Article 13 or not.
Quoting Boing Boing:
Under Article 13, anyone who allows users to communicate in public -- to post audio, video, stills, code, anything that might be copyrighted -- must be sent to a copyright enforcement algorithm that will compare it to all the known copyrighted works (anyone can add anything to the algorithm's database) and censor it if it seems to be a match.
Basically, Article 13 is going to force any service that allows content to be posted from audio to video, text and code, must be enforced using a matching algorithm akin to YouTube's Content ID program which cost $60 million and is highly inaccurate and flawed. Nobody else has ever come close.
If Google can't get it right and they've invested more into a system that is regarded as the most sophisticated there is (at present) can anyone else do better? What about smaller sites who can't afford a content matching algorithm or the costs associated with hosting it? This will drive out smaller companies if it becomes law.
Guilty until proven innocent
In a traditional legal system, the onus is on the prosecution to prove beyond all reasonable doubt that the defendant is guilty. When it comes to content matching algorithms it's guilty until proven innocent. You get fewer protections with these algorithms than you do an alleged felon.
What is this world coming to when an alleged murderer has more legal options and rights than someone uploading original content to YouTube that is being falsely accused of infringing someone else's copyright?
There is no inverse recourse
The first reaction from Google, YouTube and Facebook are to side with the complainant. There is no proper process to ensure that all claims are valid and the owner actually has the rights to the content.
In the case of false positives, if we are really going to go down this path, we should have a provision to punish someone who falsely makes a copyright claim that is proven to be invalid?
If Sony falsely claims they own a Bach recording (well 47 seconds of it) and they don't, they should get a strike against their own account or lose the ability to make an algorithm claim automatically (having to go through a manual timely process).
This is madness
If Article 13 goes through, it'll cause more problems than it aims to solve. You'll see copyright trolls laying claim to works that do not belong to them and without proper provisions for contesting an algorithm flagged match, it'll be even worse.
Governments should not be getting involved in copyright, especially when it's in the interests of private companies and media conglomerates, not the end consumer who will end being hurt and silenced by this.
If you want to stop this from happening (and time is running out) there is a site that tells you how to do it here.