On Tuesday, March 26th, 2019, the European Parliament voted in favour of the EU’s new European Union Directive on Copyright in the Digital Single Market. As I noted in Copyright, the EU and user-generated content, this has proven a controversial piece of legislation. Although intended to help protect the rights of content providers / creators, aspects of the Directive – notably Articles 3, 11 and 13 – have been seen by opponents as potentially causing damage to the Internet.
These three articles are, in summary:
- Article 3, relating to text and data mining, which could adversely hit genuine research organisations and things like tech start-ups in Europe (see Why The Copyright Directive Lacks (Artificial) Intelligence as an example).
- Article 11, (colloquially referred to as “the link tax”) which could severely restrict how we can share links, and information found on European on-line sites.
- Article 13 (the so-called “meme tax”, although its scope is far greater), which has drawn the heaviest criticism,which could do much to block or restrict the availability of user-generated content.
Article 13 – which is now, somewhat confusingly, Article 17 – has been of particular concern, as it pushes the onus of rights protection aware from rights holders and pro-actively onto the shoulders content sharing platforms, which potentially causes issues, as I noted when writing on the topic at the start of March:
Aimed at the likes of Google (including YouTube), Facebook and the like, Article 13 could fundamentally impact any platform playing host to user-generated content (UGC), including Second Life, Sansar and other virtual worlds.
Under the Article, all such services are expected to pro-actively prevent any content that might violate the Directive from being uploaded. They are to do so through the use of “proportionate content recognition technologies” – that is, automated content filtering, designed to block anything that might by in violation of copyright. However, such systems a) may not be affordable to those required to implement them, and b) don’t actually work as advertised (as is the case with Google’s multi-million-dollar ContentID system, which has been shown to be far from successful).
Drawing the loudest criticism, Articles 11 and 13 (/ 17) have seen the Directive pass through a number of re-writes, and amendments were being put forward right up until February 2019. The final Parliamentary vote was subject to intense debate, but saw the Directive pass with 348 votes in favour, 274 against. A proposal to allow MEPs to debate and vote on individual amendments was rejected by just five votes*.
So now the vote has been taken, what does this mean?
Well, contrary to the more alarmist reactions, the Internet is not about to totally implode before the weekend arrives; nor are we going to see sudden changes to policies relating to content upload within Second Life. But the risk to EU based content creators, bloggers, vloggers, streamers, etc., is still very much there, and shouldn’t be sidelined; it’s just that there are some more steps the Directive must go through.
Julia Rada, former rapporteur of the EU’s Parliament’s review of the 2001 Copyright Directive referring to the new EU Copyright Directive as “a disaster for the Internet and a generation”
The first of these will come on April 9th, when the Directive is put to the European Council for adoption. Until recently, this looked like a foregone conclusion. However, despite being one of two countries responsible for the final wording of Article 13 (/ 17) in particular, it seems Germany is having something of a change of heart, with their Minister of Justice publicly opposing the idea of upload filters:
We think #Uploadfilter is the wrong Way to go about it. Dear Union, if You want to maintain a Spark of Credibility, You support our Motion in the European Parliament. How we can prevent Upload Filters.
– German Minister of Justice Katarina Barley on Article 13 via Twitter, March 23rd
This means Germany might still retract its support, particularly giving the depth of public opposition to Article 13 (/ 17) that has been voiced. While this seems unlikely (see the video above), it is still important because five other member states – the Netherlands, Luxembourg, Poland, Italy, and Finland – have voiced public opposition to the Directive the vote. So should Germany, a key nation behind the legislation and of the EU, change its mind between now and April 9th, it would prevent a Council majority being reached. This would likely result in further negotiations n the Directive being required after the EU elections due to be held on May 26th, 2019 – elections which could see those determined to see the Directive passed over and above the many concerns raised, voted out of office.
And even if the Council pass the Directive, it will still take time to come into effect. EU nations (and probably the UK, despite Brexit), are theoretically allowed two years to transpose EU Directives into legislation; more often than not, it takes longer than this to be achieved. In the case of Article 13 (/ 17), the process could well be exacerbated by the fact – and as I again noted in my March 5th article – it contains many holes which need to be dealt with. This in itself could result in a chaotic situation – something that activists in Germany and Poland, as well as concerned organisations such as the Electronic Frontier Foundation are already taking up ahead of the April 9th EU Council meeting.
Despite potentially benefiting from the Directive and Article 13 (/ 17) Google, whose parent Alphabet has spent $100 million pursuing the development of content recognition filters, has joined with other major players within EDiMA, including Amazon, Apple, Facebook, Twitter, Snap and Microsoft, in speaking out against the vote, noting the Copyright Directive stands at odds with other European legislation:
EDiMA, representing digital businesses and online platforms in Europe, regrets the lack of clarity and practicality of two key articles of the Directive.
Specifically, Article 11 of the Directive have been found to be ambiguous and will lead to the greater likelihood of litigation. Whilst Article 13 continues to be technically unworkable because it imposes a monitoring obligation on platforms, in conflict with the e-Commerce Directive which states that there should be none.
– from EDiMA reaction: EU Copyright Directive is not
fit for [the] Digital Era
The Computer & Communications Industry Association in Europe has similarly spoken out against the passage of the Directive, referring to it as a “missed opportunity” and commenting:
We regret the adoption by the EU Parliament of the #copyright directive. We fear that upload filters and the press publishers’ right will harm on-line innovation and restrict on-line freedoms.
Some hope that gentler language added to Article 13 (/ 17) after the September 2018 vote, and which talks of platforms putting the means in place for rights holders to more ready point to infringements in order to have them removed, might present a means to prevent the more dreaded impact of the Directive. But this is very far from certain.
As such, while a significant battle against the EU Copyright Directive has been lost, it is not unreasonable to say the war is now over – but it is going to be significantly harder to win for opponents of Articles 11 and 13 (/ 17) as the fight moves to a more national level in order to persuade governments and the Council of the Directive’s flaws.
Sources:
- MEPs approve European Copyright Directive (including Article 13) – 348 votes to 274 (music:)ally).
- Tomorrow’s copyright vote explained (Julia Reda).
- The European Parliament has voted in favour of Article 13 (wired UK).
- Copyright in the Digital Single Market: European Parliament Vote (via Julia Reda – a useful voting reference in the May 2019 MEP elections).
* Ciaran Laval, via Twitter, informed me this was actually the result of an error: a group of Swedish MEPs apparently intended to vote in favour of allowing the debate on individual Articles, but pressed the wrong electronic buttons. Unfortunately, the vote stands as recorded, but the error has been noted.