The ongoing debate around the controversial Article 13 of the EU’s Copyright Directive has exposed what looks like a sizable gap between what regulators think is covered by copyright and what is actually covered by copyright. From the way Rapporteur Axel Voss and others talk about the kinds of materials they expect will be addressed by Article 13, it seems they think copyright covers only the rights to music and movies owned by major labels and production houses. They seem to think that the online services targeted by the article (online content sharing service providers or OCCSPs) will be able to “conclude” licensing agreements with all relevant rightsholders. After all, the main idea behind Article 13 is to improve rightsholders’ bargaining power in licensing negotiations with OCSSPs so that authors and artists get a bigger cut of the ad revenue those sites earn, thereby closing the so-called “value gap.”
If the world of copyright were as limited in scope as Article 13 proponents seem to think, then perhaps this licensing obligation would be achievable. However, since 1886 when the “formalities” necessary to obtain copyright protection were abolished (in Europe), the scope of material protected by copyright has ballooned to cover all original expressions “fixed” in tangible media: major studios’ movies, books, and music, yes, but also family vacation snapshots, text posts on social media, personal blogs, and so on. Here’s where the practical application of a licensing obligation starts to fall apart (and we’ll leave aside the complications around ‘incidental uses’ of copyrighted material for now) – there is just no way for any OCSSP to even identify all rightsholders whose content may appear on their site, much less enter into licensing negotiations with them.
For example, consider the popular practice of posting screenshots from one social media site to another. Even assuming that licenses for the original post were properly obtained by the first OCSSP (through agreements with each of the rights holders for the post itself as well as any depictions/reproductions of third-party content within the post), by what means could the second OCSSP (on whose site the screenshot is later posted) determine who the rightsholders are, much less find a way to obtain licenses from them? At least some proponents of Article 13 suggest that extended collective licensing (ECL) could solve this problem, pointing to a separate set of provisions (Article 9A) in the Copyright Directive.
The premise is this: OCSSPs would obtain “blanket licenses” covering all uploaded content by paying into a collective fund that would then be distributed to all relevant rightsholders, according to how often and in what ways their protected content appears on OCSSP sites. Again, this model might work if the only rightsholders in question are major studios, publishers, and labels; they already have some of the internal mechanisms necessary to track artists and distribute royalties to them. But copyright is broader than that. In reality, an Article 13-compliant blanket licensing program must either a) have a method for tracking and distributing royalties to every person on Earth whose original content appears in any way on any OCSSP operating in the EU or b) be based on a universal license fee of zero. As in, “in exchange for, and by the use of our service you hereby agree to grant us permission to make this work available, including by extending permission to other OCSSPs to make the same or similar works available on their sites.”
The former is impossible. The latter entirely defeats the stated purpose of Art. 13.
In addition to being either impossible or pointless, the ECL model for accommodating the licensing obligation of Art. 13 also ignores a subsequent provision stating that rightsholders are not obligated to license their works. A blanket licensing model would therefore need a mechanism for opting out, leaving OCSSPs back at square one for identifying and addressing the works of those rightsholders who do not wish to be part of the ECL program. Which brings up another blindspot in the logic of Art. 13 proponents: a single work may have multiple authors, multiple rightsholders. What are OCSSPs or rightsholders supposed to do when some, but not all, rightsholders for a single work agree to licensing terms? Block it.
The disparity between reality and Art. 13’s proponents’ perception of copyright is problematic on many levels.
First, if they actually believe that only the movie studios’, music labels’, and publishers’ works are protected by copyright, then their understanding of copyright is fundamentally flawed. Alternatively, it may be that major rightsholders are the only intended beneficiaries of Article 13’s efforts to enhance licensing negotiations. If true, then Art. 13 amounts to a broad value judgement about which artists deserve additional compensation to fill the “value gap” and ignores the likelihood that many independent artists will be worse off, either because it will simply be easier to block content from unknown rightsholders than to establish a compliant licensing mechanism or because of the complexities and failures of automated filtering systems (the mandatory back-up option for unlicensed works). Either way, Article 13 will lead the EU and all OCSSPs who wish to operate there toward a closed web, dominated by the works of major rightsholders and putting independent creators at a disadvantage.
Second, if Art. 13 proponents fully understand the scope of copyright and yet still believe that blanket licensing will actually translate into a bigger slice of the OCSSPs revenue for all rightsholders, then that belief can only be based on an assumption that collective licensing mechanisms can be developed to address all types of content, including a viable way to handle those rightsholders who opt out, in each of the EU’s 28 member states. Even in an over-simplified scenario where this might be possible, it seems extremely unlikely that anyone other than major rightsholders (and collective rights management organizations) would benefit from collective licensing.
Third, and perhaps most troubling, is the possibility that the licensing provisions of Art. 13 are not actually expected to result in more favorable licensing agreements for copyright holders, which implies that the impossible provisions are merely a way to drive OCCSPs toward automated filtering mechanisms, or even out of existence. We have already written about why mandating these systems is undesirable: they cannot account for exceptions and limitations to copyright, they are error prone, they only enhance the advantages of the largest OCCSPs, and the liability OCCSPs will face without them pushes operators toward even more aggressive filters. Once again, Art. 13 leads toward a more closed web (as some sites will demonstrate this week), but even worse, implementation of filtering technology for one purpose, copyright, opens the door to expanding those filters to address other policy goals.
If this all sounds bad, it is. But, Members of the European Parliament can stop it from becoming reality when they vote on the Directive next week. We urge MEPs reject the proposed text on Article 13.