The European Parliament finally adopted, on Wednesday 12 September, the draft directive supposed to adapt copyright to the digital age. It offers a clear victory to the media and cultural industries, which have fought hard for its adoption.
For those opposed to this directive – big web companies and digital freedom advocates – it is a blow to online freedom of expression. In any case, this is a hot topic: rarely has there been so much confrontation in the European Parliament and in the media, even in the hours preceding the vote.
It is still difficult to anticipate the precise effects of this directive for two reasons. Firstly, because Parliament’s version is not final: it will have to be reconciled with the version adopted by the Council of the European Union (the Member States), the services of the European Commission also taking part in the negotiations. Secondly, because it is a directive: each Member State will have to translate it into its national law and the exact outlines of the new rules may vary from country to country.
However, it is possible to see, in broad terms, what may change for Internet users if the directive is adopted as it stands.
One of the most contentious points of the directive is Article 13, which will oblige all major sites that allow Internet users to post content (social network, online video service, etc.) to enter into agreements with copyright holders (music, image, video, etc.).
This agreement (and this remuneration) will allow users to legally share copyrighted content, and Internet platforms to let them do so. If there is no agreement between the online platforms and the rights holders, the former must ensure that the protected works are not put online.
How? Difficult to say: MEPs softened Article 13, which provided filtering techniques so that platforms ensure, for each content, that an agreement with its right holders has been reached, and, if necessary, prevent their user from posting that content. The version of the article adopted on 12 September specifies only that platforms and right holders must “cooperate in good faith” without leading to an “automated blocking”, the technical filtering remaining however mentioned in the text presenting the directive. This will include all sites that allow users to post text, sound and video, with the exception of very small sites and online encyclopedias. In concrete terms, this means that almost all content posted by Internet users will be inspected and, if necessary, their publication blocked.
How will the platforms do it? The issue is not resolved: the text adopted by Parliament excludes “a general obligation of supervision”. In any case, the directive provides for an appeal mechanism in cases where a platform wrongly deletes content. On YouTube, the automatic detection of copyrighted content doesn’t satisfy anyone Internet users already know, at least partially, this system: it’s YouTube. The online video giant, owned by Google, has entered into agreements with many rights holders to share with them the advertising revenues associated with their content.
YouTube also has a sophisticated ContentID filtering system for tracking copyrighted content. Much contested, both by rights holders for its shortcomings and by users for its zeal (it frequently censors legal uses, such as parody or remix), this system could be generalized to other online platforms.
Sharing press articles
Article 11 of the Directive – the one that has generated the most lobbying – creates a “neighbouring right” for the press on the Internet. This means that when a digital platform (Google or Facebook are the best known and most directly covered by the Directive, but the Directive will apply much more widely) uses all or part of a press article, it will have to pay a sum of money, the scope and amount of which remain unknown at this stage.
This could be the case, for example, when a Facebook user shares a press article on their page: in this case, Facebook “sucks” the photograph associated with the article, its title and the first words of the article. Google News, the portal that aggregates links to press articles and recovers a portion of media content in a similar way, also seemed concerned. Example of a shared article on Facebook, which draws the title and illustration of the article. Example of the Google News home page, where article titles and, in some cases, accompanying photographs appear.
However, Parliament has added an exception, which is rather vague: when links are shared and accompanied by “individual words”. The text of the Council of the European Union is not any clearer, while leaving room for manoeuvre to the States by excluding from this right the inclusion by platforms of “non-substantial parts” of press articles.
It is difficult at this stage to say to what extent this draft directive will affect link sharing on Facebook or Google News. But if these two companies have to go to the cash register each time a user shares a link, this system could constitute an incentive to limit the place of press articles in their search results (Google) or the distribution of the latter on social platforms (Facebook).
Two recent examples of neighbouring rights for the press, abundantly documented in the scientific literature, point in this direction. In Spain, Google has stopped offering articles from the Spanish press in its Google News portal: newspaper sites have lost between 6 and 30% of their traffic (and therefore part of their advertising revenue). In Germany, Google has continued to index newspapers, but press and media giant Axel Springer noted a 7% decrease in visitors (http://www.europarl.europa.eu/RegData/etudes/STUD/2017/596810/IPOL_STU(2017)596810_EN.pdf).
Some smaller platforms or platforms that would not wish to fund newspaper publishers could deactivate the “aspiration” of content and offer their users simple links. According to some opponents, who have renamed Article 11 “link tax”, other platforms could block the sharing of links to press publishers that it does not remunerate, even if this eventuality, assure the Commission and the defenders of the directive, is not foreseen by the text.