Hello everyone

On the 23rd of June, Danish MEP Morten Løkkegaard (Danish Liberal Party / ALDE) wrote an op-ed in the Danish newspaper Berlingske Tidende defending the copyright proposal. (See https://www.b.dk/kronikker/regulering-af-internettets-vilde-vesten-er-ikke-en-begraensning-af). At the WMDK board meeting on the 25th we decided we should try to write a rebuttal and see if the paper would bring it. I'm hoping to get board approval and send it to the paper tomorrow so that it might possibly be printed or posted on their website on the 4th, the day before the vote. Obviously it's in Danish, but I've created a translation:

On the 23rd of June, Morten Løkkegaard published an op-ed defending the European Commissions controversial proposal for a new directive on copyright. His colleague in the ALDE-group, Jens Rohde, was also arguing for it on Twitter where he characterized opponents who had contacted him regarding the proposal as hackers, spammers, internet communists or perhaps all of them at once.

Løkkegaard and other supporters of the proposals have an easy foundation to argue from: Copyright is not something the average citizen thinks about in their daily life, and very few people have the time and interest to read the 33 page long proposal as well as the relevant amendments, opinions and other relevant material about the legislation, and even fewer people have the relevant knowledge to understand what the directive actually does. That makes it possible for Løkkegaard and co. to frame the debate however they like, which is why one should be observant of the fact that they never actually refer to the text of the proposal.

Contrary to Løkkegaard and other proponents of the proposal, Wikimedia Denmark, along the line of other nonprofit-organizations such as the Electronic Frontier Foundation, Creative Commons and European Digital Rights (EDRi), that the proposal is disproportionate, will strangle innovation and free knowledge excahnge on the internet, and doesn't adequately consider legal use of copyrighted material. A long line of experts have also criticized the proposal for the same reasons. We're opposed to both article 11 and 13 of the proposal, but are choosing to focus on article 13 since that is the most damaging one.

Article 13 fundamentally changes who is liable for copyright infringing material on the internet, there is no disagreement on this point. Until now, the "notice and takedown"-method has been used, where a hosting provider such as YouTube or Flickr are only liable for infringing material if they are made aware of it and then refuse to or refrain from removing it, a rational way of doing things and in line with how we traditionally look at liability. With the new directive, hosting providers have to actively enter into agreements with copyright collectives regarding the hosting providers' users' use of the rightsholders material or otherwise be directly responsible for copyright infringing material on their platforms, regardless of intent (art. 13 sec. 1 as amended). As a part of these agreements, or if the rightsholders refuse to or don't provide license agreements, the hosting provider has to implement content recognition filters that let rightsholders scan uploaded content to check whether it contains infringing content (likewise section 1).

This is where the proponents of the directive start protesting and claim that the directive does not mandate that hosting providers use content recognition filters, they're just mentioned as a possible solution, and the proponents were so nice as to remove two thirds of the mentions content recognition filters in the proposal! However, it doesn't take much consideration to realize that this is the only way to protect oneself from liability. Regardless of the wording, article 13 is a de facto mandate to implement content recognition filters. This is problematic for multiple reasons.

Content recognition filters cannot respect legal use of copyrighted material, the Danish quotation rule for example. Prior experience from platforms such as YouTube, which already uses a content recognition system has shown that when movie or game critics use short snippets of copyrighted material to contextualize their critique, something which is allowed under their respective local laws, they almost always get caught up in content recognition filters that "claim" the video on behalf of rightsholders. This is because of the automated nature of content recognition: it is impossible to manually go through the large amount of content posted on hosting providers' platforms, so rightsholders automate the process, and thus context dies. It is completely irrelevant whether your use of a text excerpt from a copyrighted book is protected under the right of quotation, an automated proces cannot adequately adress that. By the way, that's what is meant when article 13 is criticized for strangling free expression, as opposed to how Løkkegaard frames it, and it isn't changed by the weasel words regarding "fundemental rights" that has been added multiple places in the proposal after it was subject to criticism (such as recital 39 as amended).

Proponents of the proposal will probably point to article 13. sec. 2 which mandate that hosting providers establish a complaint process for users who feel that they have been incorrectly impacted by content recognition filters. Remarkably, the proposal does not further explain what such a complaint process would contain other than the fact that it should be "without delay", that rightsholders must "justify" their characterization of the content as infringing, and that the identity of users should not be revealed to the rightsholder (art 13. sec. 2). This phrasing is incredibly vague and could result in widely different protection of user rights in the individual member states, somewhat ironic for a proposal meant to harmonize copyright in the EU. Regardless of each states implementatoin, this proces essentially means that the user is guilty until proven innocent, something which is contrary to fundemental EU principle.

As a supporting organization for Wikipedia and Wikimedia, article 13 is harmful to projects like ours that freely let users change the contents of our platform, a model that doesn't work if content recognition filters have to check every contribution. Proponents of the directive have tried to protect themselves against this criticism through recital 37a which carves out hosting providers with "non-commercial purposes such as online-encyclopedias […] open source-software development platforms" or "scientific archives". This protection is inadequate.

First of all, the phrasing is incredibly vague. Unless you are literally a non-commercial online-encyclopedia, it's hard to know whether your fall under the phrasing of the carveout. Furthermore, the proposal doesn't explain how platforms that are primarily non-commercial but still have some connection to commercial enterprises should be treated. Wikimedia Commons, the media platform that supports Wikipedia, is a non-commercial venture, but the images hosted on the platform are meant for free use and thus can be used in a commercial context. GitHub is a platform that is primarily used for open source-software development and which is primarily free to use, but the platform also offers commercial solutions for companies. Are these two examples exempt from the directive?

Secondly, it's worth noting that Wikimedia and Github have spent a lot of resources making our opinion known in Brussels, so even if you assume that Wikipedia, Wikimedia Commons and projects like GitHub are covered by the exemption, it looks like it's been written specifically to appease us, but doesn't do anything further to protect smaller actors that don't have the same political impact. That is unacceptable.

The average reader would probably think that all this is extremely esoteric and hard to understand, and that is pretty much true. However, one cannot let complexity make it possible to pass what is at best good-faith legislative carelessness, at worst a deliberate limitation on the rights of citizens. Article 13, and other parts of the proposal for that matter (such as article 11), will affect anyone who uploads conent to the internet, and it is extremely important to be watchful of how the Danish members of the European Parliament vote on the proposal. The parliament votes on the directive on the 5th of July, and it is imperative that it isn't passed.

I know some of the phrasing is a bit awkward, but that's mostly down to the translation. Any feedback from the publicpolicy list would be appreciated.

-- 
Best Regards,
Matthias Smed Larsen