For those who do not have the time or inclination to read the full blog I have the most important counter-arguments summarized.
There will be no upload filter
Technically true. In the Directive the word upload filter does not appear. But Article 13 4c) calls for ‚best efforts to prevent their future uploads in accordance with „“ high industry standards „. Based on the sheer mass of data that are uploaded to platforms like YouTube and Instagram every minute, but also to small platforms such as hobby communities and forums, this is not possible to do with manpower alone. So the platforms have the choice to leave the market or introduce filters. Thus, each platform that will survive in the market will necessarily use filters in the foreseeable future.
Even proponents like Axel Voss now believe that filter, or more accurately „measures“ are necessary, as he said in an interview.
„I can not guarantee that the measures taken by the platforms to meet their liability will work one hundred percent and therefore the freedom of expression is limited sometimes.“
This statement alone should be a red flag in any democracy. Freedom of expression is one of, if not the highest good in a democracy. It should not be “ limited sometimes,“ especially not if no law is violated.
Which brings us to the second point.
The reform will not lead to censorship
Technically true. Article 13 and the related measures are not censorship. At least not directly. But they establish an infrastructure that can certainly be used for censorship. Programs that identify certain content accurately already exist, such as Shazam for music. Similarly, programs exist that examine uploaded content for copyright infringements and terms of service and take those down, e.g. the Contend-ID system from YouTube and the filter that Facebook uses to banish nudity from the platform. The new dimension required by Article 13 measures would require an complete check when uploading so that content can’t be uploaded. Combine that with a software like Siri or Alexa that can recognize spoken names and you have a filter, that keeps any criticism of a politician or a party from the Internet. And no one can guarantee that not one day a party is in the government, which does ask for exactly that function to be imlied.
YouTube, Google and Facebook are behind the protests and influence young people
Yes, YouTube has drawn attention of users to the copyright reform. But the opponents of reform are primarily digital experts (across the political spectrum), which are just now getting supported by YouTube users. A survey of 66,000 opponents of the reform revealed that 80% are over 18. Accusing all opponents, they are externally controlled and would not understand the subject, just isn’t true.
On the other hand, the collective right management organisations (CRM) attempt to influence it’s members. GEMA, a German CRM for music, sent a letter to its members in favor of the reform. 2 Supervisory Board members expressed their opinion several times in television and newspaper and tried to discredit the opponent.
YouTube however, facing the resistence, the CRM to build up with politicians, did the only thing it could do: Go directly to the voters. What’s ultimately is the fairer way to make themselves heard.
Facebook, however, is surprisingly quiet, which may be because they see an opportunity to offer their own filters as a solution for smaller platforms.
Upload filters work
The examples mentioned above like Shazam have a very high success rate to recognize music. And indeed, an upload filter with the technology would achieve relatively high hit rates. But already a deviation of the speed of a few percent would make the system useless. Also, the Content ID filter reaches high quotas what music terms. Mathematically, it is relatively easy to break down a song in a data table with the then a melody for hits is searched. But the copyright extends not only to music, although this is clearly the focus of this directive, but also pictures, films and texts. And here it starts to become difficult. A simple text filter is now used in any email client, where you can see the error rate of such seemingly simple systems, if one takes a look at what is being mistaken for spam and what passes. Coming to images, it starts to get really difficult. The similarity in possible motives and the possible variations pictures can have (mirrored, rotated, cut-outs, collages, Photoshop, just to name a few) shows how difficult it is. Let’s assume a filter is 99% is correct (100%, which will certify them IT experts, will never be reached), then every minute 4 hours of videos would be falsely claimed copyright infringing on YouTube. On Instagram that’s 400 posts, every minute. And these are the current rates. Because of the potential liability it is safe to assume rather more than too little is being blocked. And all of that does not account for the fact, that somehow the filters must be taught that exceptions, for example satire or criticism, have to pass.
The directive demands a “effective and expeditious complaint and redress mechanism“ and even legal action will be granted. But that will either be used by hardly anyone because of the cost and time involved. Or it will occupy the courts.
It’s about the artist
The aim of the policy is a fair and reasonable payment for artists. That’s a good idea. And even the opponents of the directive are in favor of this. Nobody says that platforms like YouTube should earn money without a fair share for the creatives. The resistance to the reform is not directed against the IF, but against the HOW. The reform is clearly focused on artists who belong to a CRM. That may be the right approach in principle, but ignores millions of authors, photographers, video artists, musicians, authors and designers who are not mempers of such an organization, because it does not work for them. The pattern of distribution of some collecting prefers the big artists and disadvantaged the smaller creators.
These little ones have to depend on these platforms and thus have the chance to profit from their art. Whether by advertising themselves, through jobs they acquire there, or even directly by monetization (on YouTube) or licensing (on photo-sharing sites). With the directive these small creators only have the opportunity to join an organization, which is hampered by membership fees and restrictions on membership, or to get in negotiations with the platforms themselves, where they have no bargaining power.
At the same time again granted a share of the revenue from collecting societies with Article 12 publishers who actually attributable to the authors.
With general licenses no filters are needed
Some opponents of the directive argue that it is impossible to negotiate licenses all creators of user-generatet contend (UGC). However, some proponents argue that this problem is solved by the fact that artists are not members of the CRMs nevertheless be represented these and refer to on §9a the directive. The directive provides that CRMs can negotiate licenses for those artists who are not members. So licenses be awarded without the need to negotiate with every single person on the planet.
This argument has two problems. First, each creator has the right to not be represent by a CRM. In that case, when this creator insists that his works should not be published on a platform, that platform would have to use the platform filter.
The second problem is that the licenses negotiated by the CRM are first paid to the CRMs. They have to give them to the artist. However, the directive is not sufficiently certain to ensure that this payment also goes to the artist accurately. As already described, artists that are members of CRMs are better off, but the artists who aren’t members and do not contradict see their licenses seep in the CRM and must now claim their share from the CRM. This isn’t really a progress for these artists.
Large companies will not give up their business
Twitch, one of the largest platforms for live streams has already announced to pull its business out of Europe, when the copyright reform comes into effect in its current form. Even YouTube has announced plans to close massive numbers of channels. Google had already withdrawn from China, with its search engine because it did not want to introduce the appropriate filtering mechanisms. And China is a bigger market than Europe.
A bluff from the big businesses to build pressure? Possible, but at least something to keep in mind and seriously consider before deciding on the directive. If the business is no longer profitable in Europe, even a giant like Google will consider to pull from Europe.
Small platforms are exampted
The argument that often follows the previous one, is, if YouTube, Google and Twitch disappear from the European market, then new platforms can be established. But then there are only two possible scenarios:
Either the new platforms close after three years to prevent the directive to apply to them, which would mean the CRMs won’t get anything. Or they buy the appropriate software for filtering from the big players: Google and Facebook.
Large platforms, can afford to develop such a system, but smaller platforms can’t and would be dependent on global players such as YouTube, Google and Facebook.
The users are safe
The users are only safe when the liability is transferred to the platform. An uploader, does not know if this is the case and could upload an unlicensed clip unknowingly. Should the upload happen despite the „best efforts“ of the platform, the user would still be liable. This conveys a false sense of security.
In addition to that, everyone who, for example, generates ad revenue for their YouTube Video will face the thread of no longer be covered under the YouTube license, and therefore be liable again.
Article 11, the newspapers will participate in the profits of Google
Article 11 of the directive is trying to reestablish a model of licensing to share revenues of platforms with publishers, which has already failed in Spain and Germany. In Spain, Google has simply pulled out and in Germany, the publishers have found that their bluff has been exposed, after which they have awarded licenses for free. The links on Google News were, above all anything else, one thing: advertising for the newspaper. And the same failed system is to be introduced throughout Europe with Article 11.
There is no alternative
Often it is claimed there were no alternatives. Of course there were and there still are. In the negotiations, several models were on the table. A model that the platform operator forces to take licenses to demand measures to filtering would be possible without the need for filtering. All it needs is to sit down again at the negotiating table.
It’s all scare tactics
Let’s ignore for a second, all the arguments that have been mentioned above. And let’s assume for sake of the argument, all the fears of opponents were mere scaremongering. Why does it have to be exactly this policy? Why not again come together, along with copyright and digital experts, to give the directive a little fine tuning? No one denies that a reform is necessary. And maybe all the concerns of the opponents are unfounded. Maybe we’re wrong.
But what if not?