After years of thrashing out the text of the proposed update to the EU Copyright Directive, we have come to what is almost certainly the final vote, in the European Parliament plenary early next week. You might think at this stage that it would be all over, with nothing new emerging, and most people simply accepting things as they are. Nothing could be further from the truth. The last few weeks have seen some of the most dramatic developments in the already fraught passage of the Directive through the legislative process.
For example, a few weeks ago, the European Commission published on Medium an extraordinary post entitled “The Copyright Directive: how the mob was told to save the dragon and slay the knight”. The article – now deleted, but still available on archive.org – made it clear that by “the mob” the Commission meant people who had the temerity to point out the terrible damage that the Copyright Directive will inflict on the Internet in the region. Moreover, the post claimed that “‘Big Technology’ has even ‘created’ grassroots campaigns against the Copyright Directive in order to make it look and sound as if the EU is acting against the ‘will of the people’.” In other words, not only were millions of EU citizens part of a mob, but the European Commission implied they were paid to be part of that mob – mere shills for Internet companies like Google.
The European Commission was not alone in casting aside any pretence of impartiality and honesty in discussions about the Copyright Directive. The European Parliament paid for a short video on the subject. It contains a number of deeply misleading statements: that smaller companies are “a little bit exempt” from Article 13 – a meaningless phrase – and that memes and gifs will be “protected”. Neither of those is “protected” from algorithmic upload filters, which cannot distinguish between legitimate use in memes or for parodies, and unauthorised uploads. All-in-all, the video is a disgrace, and truly unbecoming for an organisation that is supposed to represent the interests of 500 million EU citizens, not just those of a few big copyright companies. Shockingly, the video is still available on the European Parliament Twitter account, which shows that it is quite happy to disseminate this biased and misleading viewpoint.
These unprecedentedly anti-democratic interventions from the European Commission and European Parliament show beyond any doubt that supporters of the Copyright Directive are panicking – otherwise they would not feel the need to stoop so low. It’s not hard to see why. The crucial claim that Article 13 will not lead to upload filters has been demolished by a stream of weighty expert views to the contrary.
Then, we had David Kaye, the UN Special Rapporteur on freedom of expression issuing the following warning: “Article 13 of the proposed Directive appears destined to drive internet platforms toward monitoring and restriction of user-generated content even at the point of upload.” He then went on to say that “Such sweeping pressure for pre-publication filtering is neither a necessary nor proportionate response to copyright infringement online.”
The definitive confirmation that Article 13 does, indeed, mean upload filters, came from the German government, no less. In an official reply to a question submitted by a member of the country’s national parliament, Christian Lange, Parliamentary State Secretary to the German Federal Minister of Justice and Consumer Protection, stated: “In the [German] federal government’s view it appears likely that algorithmic measures will have to be taken in connection with large volumes of data for practical reasons alone.”
No one can deny it now: Article 13 means upload filters, with all the widely-recognised problems of over-blocking they bring with them. Further proof that the argument is over comes from the appearance of a new – and rather desperate – suggestion by supporters of Article 13. They are now saying that collecting societies can just give blanket licences to online companies that allow users to upload anything, which means there will be no requirement to check, and so no need for upload filters. That’s the view of Germany’s main conservative parties, the CDU and CSU, as well as of a new lobbyist organisation, “manifesto4copyright“.
If this were a serious argument, it would have been used to convince people long before – not at the last minute, when everything else has failed. If its supporters really think it should be examined as a feasible approach, then this would imply that the vote on the Copyright Directive should be postponed in order to give MEPs more than a few days to consider it properly. Since they are not calling for this, we can assume that it is simply a last-ditch effort to distract from the deep, unfixable problems of Article 13.
In any case, the suggestion is completely unworkable, as this detailed legal analysis (in German) indicates. The proposal, such as it is, assumes that suitable collecting societies exist in every EU Member State that would be capable of handling the huge volume of licensing required. Despite yet more false claims to the contrary, few digital companies will be exempt from needing a licence for material uploaded by users. Even the lead MEP for the new law, Axel Voss, admits this. In the larger EU nations, such societies may exist for music and video. But in smaller EU states, the situation is far less developed. And there are no collecting societies anywhere for material such as software, maps, choreography, 3D models etc. – all subject to copyright, and thus requiring licences, which will be impossible to obtain from everyone, leaving filters as the only option.
Finally, there’s also the issue of whether creators will be happy to hand over complete control of this kind of online dissemination of their works to collection societies. If they aren’t, and don’t wish to be represented by them in this way, then online companies will once more need filters to prevent those specific works being uploaded, so nothing has been gained by implementing this approach.
As we draw close to the finish of this long and exhausting legislative process, a few things are clear. One is that Article 13 does mean upload filters – no one can seriously deny this. Another is that Article 13 will not only harm freedom of expression on a massive scale, but it will also seriously damage the region’s already struggling digital economy. That’s the urgent message of a new open letter from 130 EU businesses. No amount of tweaking, tricks or supposed ‘fixes’ will turn this flawed technology into the long hoped-for magic pixie dust. The only solution is to drop Article 13 from the agreed text completely.
Bernd Porr is a lecturer at the University of Glasgow. He straddles the two worlds of science and the arts. He invented the fastest walking robot in the world, numerous artificial intelligence (AI) learning algorithms (that is, sets of rules), deep brain stimulation devices, and models of the emotional system of the brain. He also conducts hearing aid research, and is co-founder of a startup producing data acquisition equipment. Alongside his many-stranded academic research, Porr worked in theatre for more than 10 years as a lighting designer, wrote/produced radio programmes/radio plays, and has been making fiction films since 2008.
GM: First, could you give a quick introduction to how today’s much-hyped AI systems work in practice?
BP: AI is divided up into open loops and closed loops. The open loop is simply you have some kind of “black box” and you show the box an image. The box says: “that’s a cat”, but actually the image is a dog, so you punish the box, and hopefully the next time the box gives the right answer. If not, you punish it again. Otherwise you give the box a carrot as a reward. And that’s basically AI. Essentially you have some kind of error signal. If it’s positive, then you tell the network [inside the black box], OK, this was great, or you tell the network it was bad, and do something else. And you do this [process] in AI nowadays roughly 5 billion times. Hopefully, at the end the network does the right thing.
The closed loop stuff was what Google’s DeepMind did with Atari video games. The network looked at the video image of this very old Atari video game, and then took an action to win the game. If the network lost, it was punished. It then ran the game 5 billion times, and the network learned to solve it. Basically it was all happening in this black box, and it was becoming better and better. But it is still very error-prone, and obviously takes a long time to train. So there is no magic behind this, I would say.
GM: We know that Article 13 will inevitably lead to upload filters. The volume of digital material that will need to be filtered is beyond human capabilities, which means that upload filters must be automated. Speaking as an engineer, how would you go about using AI to achieve what Article 13 demands – the recognition and blocking of unauthorised uploads?
BP: In terms of technology, I would buy computing time somewhere in the cloud, and run the AI algorithm there. Deep learning [AI] is basically able to detect images, and [some claim] the error rate is just now around 2%. But I just had a look at more realistic articles which say the error rate is 10%.
In order to train this AI system, you need to have massive amounts of computing power. If you run this service for three years, say, Amazon would charge about £100,000 for that. You have to have more or less all the images of the world in your database. I as engineer would not know how to do that, because I would need to scavenge the Web, and then train the network on this.
Images are being created all the time, so that’s why the cost of this would easily go into the millions. For me as an engineer, I think there’s just no way of doing it, from a cost perspective, and also time. The training of the [AI] networks takes a very long time – it takes millions and millions of iterations to get to a very low error rate. Every time new films or images are generated, the training has to be updated.
The computing cost of this would be absolutely massive, and I don’t think it’s possible for anybody except Google, who already have a system [Content ID]. I think for anybody who is wanting to comply with [Article 13] they basically have to go to Google and license it. Google will have a total monopoly regarding this.
I just had a look at Amazon Web Services – they’re very advanced in offering packaged deals for image detection [in the cloud]. Even they would not be able to do [what Article 13 requires]. If they were able to do it, I’m pretty sure they would be offering the service already. So for me as an engineer, if I worked in a company, I would just say: well, we have to subcontract Google to do that. There’s no other way around this.
GM: What about the limitations of this kind of AI-based approach – for example, the well-known issue of false positives?
BP: I think that’s a major problem. The issue is that deep learning has amazing learning algorithms, but deep learning is not just one network layer, but a black box feeding into the next one – it iterates. Nobody knows what these black boxes do when they get these very diffuse rewards and punishments from the outside. So it’s very hard to figure out what decision mechanisms they use. There might be absolutely weird false positives coming up. I’m very sceptical regarding these approaches, but the problem is deep learning has been extremely hyped up, and there is very little criticism of it.
If you were a lawyer, especially in Germany, you could send what in Germany is called an Abmahnung [a demand for a payment for having found a violation of the law, for example copyright law]. You could send everybody whose site has an upload button this letter, because there will definitely be errors made in the detection protection. Even Google will make these errors.
BP: Especially in jurisdictions where people can be sued, I think that people will not allow uploading anything, any more, because of the [legal] risk. So I think the danger is that basically a lot of smaller companies will just block completely any upload. For me, the main worry is that upload filter systems will default to the assumption that you are a criminal if you are uploading anything there. I think it will cause mayhem.
Anriette Esterhuysen was the executive director of the Association for Progressive Communications (APC) until March 2017. Prior to joining APC, Anriette was executive director of SANGONeT, an Internet service provider and training institution for civil society, labour and community organisations. She was active in the struggle against Apartheid from 1980 onwards. From 1987 to 1992 she was involved in information and communication work in development and human rights organisations in South Africa and Zimbabwe.
Recently, she was one of the signatories of an open letter to the President of the European Parliament, asking him to delete upload filters (Article 13) from the EU’s proposed Copyright Directive, formally known as the Directive for Copyright in the Digital Single Market (DSM).
GM: When did you first use the Internet?
AE: Pretty early on. I first used email in 1987 and online discussion forums in the early 1990s. These systems evolved into what we currently know as the Internet.
GM: When did you first become aware of copyright? What did you think of it?
AE: I worked in university libraries in the late 1980s and also worked as a sociology tutor. This made me aware of the huge restrictions placed on university libraries when it came to photocopying material for students who could not afford text books.
But my awareness increased when I started working on providing online content. I learnt that something really simple, like translating training material into a language that the people you were training could understand, constituted a copyright evaluation. Or sharing educational resources. I became part of the Creative Commons movement, and worked with a team of local pro bono lawyers to establish a South African Creative Commons License.
For me the most striking absurdity remains the fact that school kids in rural areas in my country, South Africa, have to go through their school year without text books, because the text books can only be accessed in hard copy, with the publishers of the text books not allowing any form of online distribution because of copyright restrictions.
As the Internet became ubiquitous, the increasingly desperate measures of publishers, and the music industry, convinced me that the potential of the Internet to democratise access to information was at risk. Usually it was not the artists themselves who wanted to increase copyright enforcement, but the intermediaries – the distributors – who in most cases were the beneficiaries of a business model which involved exploiting both creators and users to control sharing of content over the Internet.
I realised that copyright, while a good thing in principle, was being distorted and abused by corporations who pretended to be concerned about the interests of creators. But all they really cared about was loss of profit. They dominated the policy debate, and much-needed exploration into new business models for content distribution, and open licensing systems, was not happening.
GM: Could you say a little about what the Association for Progressive Communications is, and what it is trying to achieve?
AE: The Association for Progressive Communications is global membership-based network of organisations and individuals, established in 1990, who work collaboratively on the Internet and information and communication technologies (ICTs) for social justice, gender equality and sustainable development. APC has 58 organisational members and 34 individual members in 74 countries, primarily in the global South. Access to information and using ICTs to strengthen the global commons have been important to APC since our establishment. We have advocated for open licensing for more than 20 years.
GM: What role does the Internet play?
AE: It is a disruptor and an enabler. What is disturbing is that the same technologies that initially opened the doors to content sharing, and access to information to people who cannot afford to pay for it, are now being used to reinforce digital rights. It is also disturbing that policy makers – including in Europe – are spending more time listening to those with vested interests in reinforcing intellectual property rights than encouraging the creation of new business models that will benefit both creators and users.
GM: How does copyright affect APC’s work?
AE: The ease of copying and sharing information challenges conventional notions of copyright control as it evolved in the late 20th century. But we should not forget that this system of control was one that was developed by ‘middlemen’: publishers and an entertainment industry – not by individual authors or creators.
GM: Moving on to the proposed EU Copyright Directive, you signed the letter about Article 13: what do you see as the problems there?
AE: It starts from the wrong place. This is why I think that even though some of the improvements in the latter half of 2018 should be noted, the whole initiative is still problematic. The main problems:
1) It changes the liability regimes by making intermediaries responsible for copyright enforcement, proactively. This puts freedom of expression at risk, but it also puts the right and ability to access information of millions of users at risk. The current system of following up on perceived violations gives users the ability to make a case. Automated systems won’t.
Perhaps there is a need to look at liability in the context of a platform like, e.g., Facebook, as they are clearly not just carrying content, but also manipulating who sees what. But that is a different issue. Making intermediaries liable for the content they carry will a) undermine freedom of expression and b) prevent millions of users who cannot afford to pay systems and who do not have credit cards from accessing information in the first place.
2) Definitions are vague. The following definition [from the Copyright Directive] applies to many not-for-profit entities that disseminate content online: “Information society service providers that store and provide to the public access to large amounts of works or other subject-matter uploaded by their users”.
3) It starts from the wrong place. If the proposals started with user rights at its core, then there would be some definition of fair use (which there is still not, as far as I know) and then provisions would be built on top of their rights, rather than slashing through them, and then trying to add fixes after the fact.
GM: What do you think should be done about these problems?
AE: Go back to the drawing board. There is definitely a need to look at the need of journalists, and of news outlets, to be financially sustainable on the Internet. Also of other creators of content. But this process should start with all stakeholders around the table, with both the interests of rights holders and users being considered at the same time. It should explore new business models and approaches to copyright rather than try to maintain models that are clearly no longer viable through top-down controls.
GM: What about other parts of the Directive – for example, Article 11’s ancillary copyright: what’s your view on that?
AE: I think the so-called “link tax” is a disaster. Perhaps its intention is to strengthen mainstream journalism and media – but that is not likely to be the outcome. Having to pay for quoting news sources is likely to strengthen fake news rather than increase the income of the news sources.
It is very conceivable that platforms and search engines will enter into contractual agreements with large media outlets, allowing their content to be quoted free of charge…thereby side-lining smaller and independent media outlets and undermining diversity in the media environment as a whole. Again, the impact for the millions of users (including journalists and creators of content, researchers and activists) who cannot afford to pay will be massive. Think of an Internet divided into a ‘have money, can pay, will quote legally’ part, and a ‘do not have money, or credit card, cannot quote legally’ part? It seems insane to me, particularly if your goal is to get your information out there and used.
GM: Any thoughts on Article 3, which covers text and data mining?
AE: I think the intentions are good, but I can’t see it having good outcomes. It could limit the use of data by academic researchers, and by civil society organisations. And I might be wrong, but I don’t see it impacting on the huge Internet companies and the data brokers they work with, who are the biggest data miners.
GM: If the Copyright Directive is passed in its present form, what do you think its knock-on effects around the world might be? For example, what effects might it have on access to information, and on social justice and development around the world?
AE: It will restrict access to information by people who need it, are using it, and who cannot afford to pay for it. I see it as fundamentally undermining the ability of the Internet to create a more level playing field in terms of access and creation of information.
GM: South Africa has also been grappling with updating copyright law, and there is heavy lobbying around this. What has been happening there, and how would you contrast it with what is happening in the EU?
AE: The amendment bill was passed last year. The recognition of fair use is a great step forward. I think that in many respects, while not perfect, the South African copyright amendment process was inclusive and consultative, and did consider the rights of users in spite of heavy lobbying from the entertainment industry. The bill also addresses exceptions and limitations for users with disability.
How to contrast it with what is happening in the EU…? Overall, I think it was a more user-centred process. This is how it should be. Information is so vital to development, democratisation, protection and promotion of rights. Any effort to reform copyright should use that as a starting point. Not the interests of industries who are resisting fundamental change to their current business models.
In a previous editorial, I pointed out that at the heart of Article 13 in the proposed EU Copyright Directive there’s a great lie: that it is possible to check for unauthorised uploads of material without inspecting every single file. The EU has ended up in this absurd position because it knows that many MEPs would reject the idea of imposing a general monitoring obligation on online services – not least because the e-Commerce Directive explicitly forbids it. Instead, the text of Article 13 simply pretends that technical alternatives can be found, without specifying them. The recently-issued “Q and A on the draft digital copyright directive” from the European Parliament then goes on to explain that if services aren’t clever enough to come up with other ways, and use upload filters, then obviously it’s their own fault.
Imposing legal obligations that are impossible to fulfil is deeply irresponsible law-making. But there is another aspect of Article 13 that is even worse: the fact that it will encourage a new wave of criminality. It’s hard to think of a greater failure than a law that increases lawlessness.
The problem arises once more from the flawed idea of forcing companies to install upload filters. Just as EU lawmakers seem unable to grasp the fact that online services will be obliged to conduct general monitoring in order to comply with Article 13, so their lack of technical knowledge means that they don’t understand the tremendous practical challenges of implementing this form of general monitoring.
At least the French government is much more consistent and honest in this regard. It wants to go even further than the agreement it struck with the German government, which ended-up forming the basis for the Article 13 text in the Romanian Council Presidency’s new mandate that was adopted on Friday, 8 February. France wants to remove the references to Article 15 of the e-Commerce Directive, which prohibits Member States from imposing general monitoring obligations, in order to ‘clarify’ that these types of obligations are perfectly fine when it comes to protecting copyrighted material.
Another editorial pointed out some of practical challenges of implementing this form of general monitoring. Article 13 will apply to every possible medium. That means online services will need filters for text, music, audio, images, maps, diagrams, photos, video, film, software, 3D models etc. Material can only be filtered if there is a list of items that must be blocked. So in practice, Article 13 means that any major site accepting user uploads must have blocking lists for every kind of material. Even where they exist, such lists are incomplete. For many areas – photos, maps, software, etc. – they simply don’t exist. Moreover, for much of the content that would have to be monitored, the filters don’t exist either. In another instance of irresponsible, lazy law-making, Article 13 is asking for the impossible.
What will online services do in this situation? The Copyright Directive doesn’t help – it simply says what has to be done, not how to do it. This will incentivise companies to set up systems that are likely to provide the best protection when it comes to the inevitable lawsuits. The primary concern will be blocking with maximum efficiency material that should be blocked, rather than on minimally intrusive approaches that maximise freedom of speech for users. The lack of systems to protect themselves from this liability could also mean that some platforms will use geo-blocking, just disappear, or move away from the EU, and that others will never be created in the first place in Europe.
This emphasis will encourage the setting up of systems that allow anyone to submit claims on material, which will thereafter be blocked. Adopting this system, companies will be able to handle material for which general blocking lists do not exist, and will thus be able to avoid liability for unauthorised uploads. As well as being the only practical way to handle the enormous challenge of filtering every kind of copyright material, it also has the virtue of being an approach that has already been used elsewhere, albeit on a smaller scale.
For example, YouTube allows anyone to claim that they are the copyright holders for material that has been posted to Google’s service, and to cause it to be removed automatically. The negative consequences of that feature were discussed previously; suffice it to say that legitimate material is often taken down by mistake, and that appealing against those decisions is hard and time consuming, and the results are very unpredictable. The same will inevitably happen with Article 13’s upload filters, with the added twist that material will be blocked even before it is posted, whereas the automated takedown system created by the US Digital Millennium Copyright Act (DMCA) only operates after material is posted online. However, a recent story on TorrentFreak reveals another disturbing possibility:
In a terrible abuse of YouTube’s copyright system, a YouTuber is reporting that scammers are using the platform’s “three strike” system for extortion. After filing two false claims against ObbyRaidz, the scammers contacted him demanding cash to avoid a third – and the termination of his channel.
Under Article 13, it doesn’t even require three strikes: if your upload gets caught by the filter, your material will be blocked forever. The thinking seems to be that it doesn’t matter if mistakes are made, because people can simply appeal. But as we’ve noted before, appeals processes are slow, don’t work and aren’t used by ordinary people, who are intimidated by the whole process. So even the threat to claim material will be much more powerful with Article 13 than it is with YouTube.
This means that no one can guarantee that their content will make it online in the first place, except for the big (US) rightholders who will have forced the major (US) platforms into agreeing to licencing deals. If your content does actually make it past the upload filter, then you will still run the risk of being extorted by copyright scammers abusing the system. The stay-down obligations of Article 13, implying that copyrighted material that has been flagged by rightholders (or scammers) can never be re-uploaded, makes attempts to claim material or get something back online even harder under Article 13 than they currently are with YouTube.
This is particularly bad news for new artists, who desperately need exposure, and don’t have deep pockets to pay for lawyers to deal with these kind of problems, or spare time to do it themselves. More established artists will lose revenue all the time their material is blocked, so they too may decide to pay up to scammers issuing false copyright claims. This new threat will also see activists’ use of sites allowing public uploads badly affected: many online campaigns are tied to particular events or days. They lose most of their power if they are delayed by weeks or even days, which appeals processes will certainly involve. Far simpler to pay the blackmailer.
This problem exposes a further flaw in Article 13: there are no penalties for falsely claiming to be the copyright holder for material, and causing legitimate uploads to be blocked by upload filters. That means there are almost no disincentive to sending out thousands – maybe millions – of automated threats to artists, activists and others. It’s true that this is extortion, and thus illegal. But given how overwhelmed police forces are today, is it likely that they will allocate scarce resources to chase down phantoms across the Internet? It’s easy for people to hide behind fake names and temporary accounts, and to use anonymous payment systems like Bitcoin. With enough time and effort it might be possible to establish who is behind them, but if the sums demanded are low, the authorities simply won’t bother.
In other words, the poorly thought-out nature of Article 13’s upload filters risks creating a new class of “perfect” online crime. It’s one that can be conducted by anyone with an Internet connection, from anywhere in the world, and one that is practically risk-free – a seductive and deadly combination. Far from helping artists, the Copyright Directive could create a massive new obstacle to their success.
Both the Romanian Council Presidency text adopted on Friday 8 February by COREPER I and the version of the copyright Directive adopted in the European Parliament in September 2018, which will form the basis of the trilogue starting on Monday evening, 11 February at 20h CET, have been met by an increasing opposition from a wide-variety of stakeholders, both as regards Article 11 on the press publishers’ right, article 13 on upload filtering and/or even against the Directive as a whole, and this for a number of sometimes diverging reasons.
The overview below of those who have voiced their concerns can only leave one with the feeling that something must be utterly wrong with the current proposals on the table, if all these stakeholders– having had opposing views throughout the debate – start to agree on the same thing, namely: that everyone would be better off without the proposed reform. It also indicates that there is no single voice representing creators or journalists, whilst there has been a consistent and unified view expressed from the users’ side to say no to Article 11 and 13.
This should give a strong signal to policymakers that moving forward with the copyright reform in its current form is the wrong option, and that the best legacy this European Parliament could leave us would be to decide to reject the whole copyright proposal that is on the table. Alternatively, the institution representatives at the trilogue this week could simply delete Articles 11 and 13, which would go a long way into ensuring the Copyright Directive finally becomes more balanced.
Recent Positions by Stakeholders
8 Feb, 2019
IFJ –International Federation of Journalists (representing 600.000 media professionals)
EFJ –European Federation of Journalists (representing over 320.000 journalists in 70 journalists’ organisations across 44 countries)
Broad community of stakeholders ranging from civil society organisations, creators, academics, universities, public libraries, research organisations and libraries, startups, software developers, business organisations, EU online platforms, to Internet Service Providers (+85 Organisations), including:
European Digital Rights (EDRi)
Civil Liberties Union for Europe (Liberties)
Copyright for Creativity (C4C)
European Bureau of Library, Information and Documentation Associations (EBLIDA)
European Internet Services Providers Association (EuroISPA)
European University Association (EUA)
Ligue des Bibliothèques Européennes de Recherche – Association of European Research Libraries (LIBER)
Scholarly Publishing and Academic Resources Coalition Europe (SPARC Europe)
Association for Progressive Communications (APC)
Electronic Frontier Foundation (EFF)
Electronic Information for Libraries (EIFL)
Index on Censorship
International Federation of Library Associations and Institutions (IFLA)
The Article 13 cliffhanger continues. Disagreements between France and Germany over exemptions from the requirement to use upload filters halted progress on finalising the new legislation, and offered hope Article 13’s deep damage to the Internet might be avoided at the last minute. But the two countries appear to have reached a compromise that is arguably worse than the original text. It implies that in practice, even the smallest sites will be forced to seek licences, and accept whatever terms are offered to them. This is a recipe for even more abuse from the copyright industry, and will drive digital startups away from the EU.
But alongside this incredibly foolish proposal from France and Germany, there’s an interesting comment from Luigi Di Maio, the Italian Deputy Prime Minister and Minister of Economic Development (original in Italian), which has been largely overlooked:
The priority is modifying Articles 11 and 13, which deal with the link tax and content filtering. These are hectic days on the Copyright Directive front. The signs coming to us from Brussels are not encouraging, but I am confident one can find a solution that protects the rights of Web users while guaranteeing at the same time the rights of authors.
What’s significant there is the mention of users’ rights. Discussions about them have been conspicuous by their absence most of the time the EU’s Copyright Directive has been under development. That is truly scandalous, and highlights just how one-sided the proposed legislation is. It is all about giving yet more rights to the copyright industry, with no regard for the negative impact on everyone else. That overriding consideration is so extreme that the dire consequences Article 13 will have on the Internet in the EU were first denied, and then ignored.
Currently, Article 5 of the 2001 Copyright Directive says that “Member States may provide for exceptions or limitations”, including “for the purpose of caricature, parody or pastiche”, which could cover memes, depending on how a judge interpreted it when it came to a court case. If EU politicians cared even the slightest about ordinary users of the Internet, the least they would do would be to make those exceptions mandatory so as to provide a well-defined legal space for memes. This is precisely what MEP Reda proposed in her 2015 report for the European Parliament evaluating the current 2001 Copyright Directive. She wrote:
The parody, caricature & pastiche exception should apply regardless of the derivative work’s purpose. It should not be constrained by a right holder’s copyright, but only by the moral rights of the author.
She also proposed a much broader recognition of users’ rights, allowing them to exploit digital technology, particularly mobile phones, to create new works based on elements from their daily lives – photos, videos, and audio – together with material they encounter on the Internet:
Copyright legislation should not stand in the way of this unprecedented wave of emerging creative expression and should recognise new creators as valid cultural actors and stakeholders.
A truly modern copyright law would embrace this exciting new dimension. For example, Article 29.21 of the Canadian Copyright Act has a wide-ranging exception for user-generated content. Its existence demonstrates that the inclusion of something similar in EU law is not an unreasonable request, and is compatible with international treaties governing copyright. And yet the proposed Copyright Directive ignores this aspect completely – and with it, the needs and aspirations of hundreds of millions of EU citizens whose lives have been enriched through online self-expression. Instead, there has been little more than lip-service to the concerns of this key stakeholder group. Here, for example, in a recent “non-paper” – the very name betrays its marginal nature – the European Commission suggests a tiny concession for users:
the co-legislators could provide that minor uses of content by amateur uploaders should not be automatically blocked… nor trigger the liability of the uploader.
But there is no explanation of how that will happen – by magic, perhaps. Instead of these vague words, we need a concrete exception that recognises the reality of how most people use the Internet these days – to share elements of copyright material for non-commercial purposes, for the entertainment and edification of family and friends.
If it’s too much to hope for a full and proper exception for user-generated content that forward-thinking countries like Canada have introduced, there is an alternative that even timorous lawmakers should be able to accept. Article 10.2 of the Berne Convention, the overarching framework for copyright laws around the world, reads as follows:
It shall be a matter for legislation in the countries of the Union, and for special agreements existing or to be concluded between them, to permit the utilization, to the extent justified by the purpose, of literary or artistic works by way of illustration in publications, broadcasts or sound or visual recordings for teaching, provided such utilization is compatible with fair practice.
Why not create a general exception to EU copyright for such “illustrations”, which would apply beyond educational establishments, to general members of the public re-using materials for the limited purpose of “illustrating” a thought or a comment? After all, it could be argued that such a use is, indeed, a new kind of “teaching”, in the sense that it imparts knowledge and opinions about the world, drawing on the possibilities opened up by modern technology. It’s not the best solution, but it’s better than nothing. It would at least show that the European Commission, Member States, and MEPs are aware of the public’s existence, and are prepared to throw a tiny crumb in its general direction.
In fact, for a short while recently, the proposed text for the Copyright Directive included such language in a section on user-generated content. This was initially proposed under the Austrian Presidency in December 2018, asked for by the Germans in their non-paper from January 2019, and initially picked-up by the Romanian Presidency. However, the Romanian Presidency then took it out after complaints from some EU countries (most likely the French). Maybe Italy should get it put back again.
In the never-ending dossier on copyright reform, the latest development makes one wonder if we finally reached the fourth dimension. Following two weeks of silence and closed door negotiations between France and Germany, the Romanian Presidency came out with its new proposal for a negotiation mandate to be voted on this Friday, 8 February 2019, at the Council’s COREPER I meeting, which brings together the Deputy Permanent Representatives to the EU of the Member States (= the deputy Ambassadors), as point 27 on the agenda (close your eyes: it’s Friday after 6 pm and when everyone is already half headed home for the weekend, they get to hit the golden buzzer for the copyright Directive to go to the final trilogue).
Now that we have the setting of this tragedy, let’s explore its plot, as that’s where that fourth dimension element kicks in.
Part 1 of the plot: The Germans cave-in
First, there is the fact that a German-French negotiation started under the pretence of Germany wanting to protect small enterprises from the burden of Article 13 – aka the upload filter, aka the #CensorshipMachine, aka welcome to the licensing utopia – by creating a carve out for them, a sentiment apparently shared by the European Parliament when they adopted their own negotiation mandate at the September 2018 plenary. ‘Protecting SMEs is a red line’, German diplomats were quoted as saying (albeit without specific reference to their name). But journalists forgot to ask them the shape and nature of that line, as the compromise reached with the French clearly show the line was dotted, curved, and easily erased.
The end res ult as proposed by the Romanians is an unrealistic mess at a variety of levels in terms of what it entails:
an SME carve-out that is not a true carve-out: The text put forward maintains the notion that licensing is the magical solution, and that everyone, be it small or large companies, is in a position to obtain licenses. This raises issues at multiple levels: (1) there are no licensing mechanisms for all the content that would fall under the scope of Article 13, and (2) for those types of content that can be licensed, this approach risks favouring an exclusive dialogue between the big (US) platforms and big (US) rightholders, excluding the smaller platforms and creators. This would severely jeopardise the EU’s cultural diversity and economic potential for both the entrepreneurial and creative sector.
a solution that will discourage EU startups, and kill EU scaleups: The EU desperately wants to foster an EU startup economy, but is simultaneously creating a regulatory environment that is discouraging these companies to come to existence in the first place. Moreover, the text put forward is killing any potential ambition for startups to scaleup, as the supposed SME-carve-out – which is not a true carve-out (see point 1) – sends a signal that the EU prefers to have small, unsuccessful companies that do not innovate or support the EU job and growth creation, whilst creating an environment that enables the existing (US) giants to thrive at their and our expense. Check out Techdirt’s specific analysis here. Finally, remember that any ambitious startup that could benefit from the criteria put forward today for the partial carve-out will lose that benefit by the time this legislation is transposed, if one considers an average of 2 years for implementation.
a UGC exception that does not even cover ‘memes’: The deletion of reference to the ‘illustration’ exception renders the proposed user-generated content (UGC) exception useless to protect harmless everyday practices by users, such as ‘memes’, as these would not necessarily fall under the other limited set of exceptions that are covered, such as parody (often defined in a limited manner) or quotation (which does not cover audiovisual material in many Member States).
an obligation to prevent the availability of unauthorised content and to ensure that this content stays-down: These obligations can only be practically fulfilled by platforms through the adoption of automated filtering technologies. Although avoiding explicitly mandating these systems, the requirements laid upon platforms leave them no other choice than to implement, or even create, such filters. Moreover, a stay-down obligation would equate to a general monitoring obligation, which would be contrary to Article 15 of the e-Commerce Directive, since a platform can only prevent re-uploads of any works by filtering all works.
So that’s on the Council side. Add to that the bizarre statements of Rapporteur Axel Voss of the European Parliament yesterday, 5 February 2019, that “What France and Germany have agreed is a new safe harbour for small platforms, able to exist beyond what we already have today. It’s something we can’t accept,” and just reach the conclusion we did: all German elements involved in this dossier just buried any hope for SMEs to ever survive this provision, and for anyone but the rightholders’ interests to be defended.
Part 2 of the plot: The Italians Step Up
In parallel to these shady backdoor negotiations (and one cannot help wonder what the Germans received from the French in exchange for their submission), Italy yesterday once more stepped forward, not just to denounce the failures of Articles 11 and 13, but to actually outline what is needed to make them realistic, workable and beneficial to creators without wrecking the entire Internet.
In a statement made on 6 February, Luigi Di Maio, the Italian Deputy Prime Minister and Minister of Economic Development, talks about such ‘crazy’ concepts as protecting the rights of web users and avoiding filters, direct or indirect, as an outcome of Article 13.
Let’s now hope that contrary to Shakespearian tragedy plots, this one does not end with an element of tragic waste or an absence of poetic justice, as EU citizens and creators deserve a better deal when it comes to copyright.
Brussels is often accused of being a bubble. Usually, such claims are made at the Member State level, by those keen to portray the European Union as being out-of-touch and uncaring.
However, it could also fairly be made as concerns the way in which law-makers take account of what is going on elsewhere in the world. Other than frequent assertions about American platforms, there has sadly been little curiosity about the wider world.
This matters. Europe is in a global competition. It owes it to its citizens to perform, by supporting innovation and productivity, today and in the future. Looking at the decisions being taken elsewhere – and the principles that underpin them – is essential.
This article therefore looks to Singapore, a country which ranks in the top three globally for education (according to the OECD’s PISA study) and top globally for innovation inputs in the Global Innovation Index. Given its influential position in Asia, where Singapore goes others are likely to follow.
Singapore launched a consultation into its own copyright laws in autumn 2016, soon after the Commission published the draft Directive on Copyright in the Digital Single Market. It received many replies from all sectors, including creators, publishers, technology companies, libraries and businesses. After two years of reflection, it has now published a report which reflects on input received, and sets out plans for next steps.
The process – and the results – contrast strongly with the situation in Europe in three ways which will be set out below: 1) the focus on the public interest; 2) a concentration on the practical, rather than the theoretical; and 3) an emphasis on simplifying the law to the benefit of users.
Putting the Public Interest First
Singapore’s focus on public interest goals is clear in the approach it has taken to text and data mining. While the European Union is still wrangling over whether to allow companies with legal access to copyrighted works to carry out text and data mining, Singapore has made it clear that the right to read should be the right to mine.
The logic is clear. Text and data mining no more competes with sales of original works than selling paintings competes with sales of paint. Indeed, it can increase demand by allowing users to do more with what they buy or access.
The same goes for the proposed rules to bring Singapore in line with the Marrakesh Treaty. These, notably, include a rejection of the supplementary remuneration provisions which the relevant EU legislation allows Member States to introduce or maintain. Given that such provisions represent a tax on the libraries, charities and others serving people with print disabilities, Singapore’s position is to be applauded.
Caring about What Works
Singapore has also shown its readiness to reject ideas which may work in theory, but which in practice are likely to prove unfair or impractical. In its proposals on education for example, the government suggests that where works are already available online, then it should be possible for teachers to use them in their work. The government rejects proposals to introduce licensing schemes, underlining that collective management organisations tend to have very poor coverage of creators online.
Similarly, it has acknowledged clearly that where other uses do not conflict with original markets – for example preventative copying in libraries, or copying for other internal uses – there should be no question about an exception applying.
This contrasts with the approach in the European Union, which leaves a broad margin for the use of the licensing schemes so clearly excluded by Singapore’s government, and leaves cultural heritage institutions still facing a lack of clarity as to whether they can make internal copies.
Making Things Simple
Finally, Singapore has also focused on simplicity, raising the chances of its laws being consistently and easily applied. In addition to the text and data mining exception above, it has ensured that all library, archive and museum exceptions cannot be overridden by contract terms, and are technology neutral.
Singapore, had already realised the importance of implementing a ‘fair use’ provision as a means of ensuring that copyright both protects the interests of creators and promotes innovation at a time of technological change. The government is now proposing to remove the fifth factor of its fair use test, making it easier to apply. For those who are not familiar with this test, the five factors (of which only the first 4 will remain in place) are:
The purpose and character of the use, including whether the use is commercial in nature or for non-profit educational purposes;
The nature of the creative work;
The amount of the creative work that has been copied, or whether the part that is copied is substantial to the whole of the creative work;
The effect of the use on the potential market for, or value of, the creative work; and,
The possibility of obtaining the creative work within a reasonable time at an ordinary commercial price.
Singapore Copyright Review Report – pp. 25-26
It is worth noting, of course, the other proposed reforms referenced above which promise to benefit teachers, innovators and those working to preserve and give access to heritage.
Meanwhile, the EU has created complexity through proposing rules only for digital (rather than physical) uses, through inconsistent application of measures on contract override, and of course through Articles 11 and 13 which, in their current state, look set to strengthen the hand of those who already enjoy a powerful position in the market.
Clearly not all is perfect with the proposals made in Singapore. For example, provisions around banning the sale of set-top boxes which are specifically designed to allow for signal piracy may risk taking with it legitimate products. Nonetheless, this is a far more proportionate and targeted approach to dealing with piracy than those currently promoted in the EU.
What does stand out from Singapore’s proposals is the readiness to weigh up the arguments on the table, and resist those that are impractical, overly complex, or do not serve the public interest. As the government’s report underlines, when creators shared their views, they tended to favour making access and use easier.
It is hopefully not too late for the EU and its Member States to learn from this example.
With the EU copyright negotiations having reached a crucial stage, over 85 organisations, representing a broad community of stakeholders ranging from civil society organisations, creators, academics, universities, public libraries, research organisations and libraries, startups, software developers, business organisations, EU online platforms, to Internet Service Providers, are sending a strong signal to the EU legislators: delete Articles 11 and 13 from the copyright reform proposal!
On 29 January 2019, over 85 organisations sent an open letter [PDF] on the EU copyright Directive in the Digital Single Market to the EU Member States’ Deputy Ambassadors, European Commission Vice-President for the Digital Single Market Andrus Ansip, and the Members of the European Parliament (MEPs) involved in the trilogue negotiations with the Council.
This initiative follows the recent deadlock that was reached at the Council level on 18 January at the meeting of the Committee of Permanent Representatives (COREPER 1), at which the EU Member States’ Deputy Ambassadors discussed the proposals put forward by the Romanian Council Presidency in order to continue their negotiations with the European Parliament (EP) with the goal of concluding a trilogue agreement. However, a number of EU Member States opposed the direction that the Romanian Presidency wanted to take, and together these Member States reached the threshold to form a so-called ‘blocking minority’, as together they represented more than the minimum required 35.01% of the EU population to stop this moving forward. The result was that the trilogue meeting between the Council and the EP, scheduled for 21 January and which had already been proclaimed as the ‘final’ one, had to be postponed.
This letter thus comes at a crucial moment in the legislative discussions, as it gives a strong signal to all EU policymakers that the best way forward out of this situation is to delete Articles 11 and 13 from the copyright reform proposal.
We can only hope that the legislators will follow this message and stop the horse trading around these provisions that, if adopted, will be detrimental to both creators and citizens’ freedoms.
The EU Copyright Directive is now in the last stages of its passage through the EU’s legislative system. Given the advanced nature of the discussions, it is therefore highly surprising that the Legal Affairs (JURI) Committee responsible for steering it through the European Parliament has recently published a “Q and A on the draft digital copyright directive“. But it’s not hard to guess why this document has been published at this time. More and more people are waking up to the fact that the Copyright Directive in general, and Article 13 in particular, will cause huge harm to the Internet in the EU. This Q and A is an attempt to counter the objections being raised, and to quell the growing calls for Article 13 to be scrapped.
The first question of the Q & A – “What is the Copyright Directive about?” – underlines the core problem of the proposed legislation. The response is as follows: “The proposed ‘Directive on copyright in the Digital Single Market’ seeks to ensure that artists (especially small ones, for example musicians), and news publishers and journalists benefit from the online world and the internet as they do from the offline world.”
Nowhere is there any mention of the EU citizens who use the Internet, or of their priorities. So it is not surprising that the harm the Copyright Directive will cause hundreds of millions of Internet users is never addressed – because the Copyright Directive’s backers simply don’t care. The Q & A claims: “What is currently legal and permitted to share will remain legal and permitted to share.” Although that may be correct in a literal sense, Article 13’s requirement for upload filters means that in practice it is far from the truth. Material that is perfectly legal and permitted to share will be blocked by the filters, because they are necessarily imperfect, and because companies faced with legal consequences will always err on the side of caution and over-block.
The next question is: “How will the Directive affect ordinary users?” The answer is, again, true but highly misleading: “The draft directive does not target the ordinary user.” Nobody is saying that it targets Internet users – in fact, they are totally ignored in the legislation. But the key point is that upload filters will hit ordinary users, and very hard. Whether or not this is the intention is irrelevant.
“Will the directive affect internet freedom or lead to internet censorship?” the Q & A asks. The answer here is: “a user will be able to continue uploading content to internet platforms and (…) these platforms/news aggregators will be able to continue hosting such uploads, as long as the platforms respect the creators’ right to fair remuneration.” Yes, users can upload content, but some of it will be blocked unjustifiably because platforms will not take chances on material that may not be covered by any licensing deals they have signed.
The next question touches on the great lie at the heart of the Copyright Directive – that it does not require upload filters. This is something supporters have been promulgating for some time, and it is shameful to see the European Parliament itself repeat this untruth. Here’s the key part of the answer:
“The draft directive sets a goal to be achieved – An online platform/news aggregator must not earn money from material created by people without compensating them. Therefore, a platform/news aggregator is legally liable if there is content on its site for which it has not properly paid the creator. This means that those whose work is used illegally can sue the platform/news aggregator.The draft directive however does not specify or list what tools, human resources or infrastructure may be needed to prevent unremunerated material appearing on the site. There is therefore no requirement for upload filters.
However, if large platforms/news aggregators do not come up with any innovative solutions, they may end up opting for filters.”
The Q and A is trying to claim that it doesn’t require upload filters, and that the onus is on Internet companies to come up with “innovative solutions”. It says plainly that if a company uses upload filters, it must be to blame for not being sufficiently “innovative”. This is deluded nonsense. Countless experts have pointed out that it is impossible to “prevent unremunerated material appearing on the site” unless every file is checked, and then blocked if necessary – an upload filter. No amount of “innovation” will get around the logical impossibility of complying with the Copyright Directive’s goal without using upload filters.
As well as being irresponsible law making, this approach also exposes the profound tech illiteracy of many EU politicians. They evidently still think that technology is some kind of magic pixie dust that can be sprinkled on problems to make them go away. They have little understanding of the digital field, and yet are arrogant enough to ignore the world’s top experts in the field, when they say that what the Copyright Directive demands is impossible.
Adding insult to injury, the response to the question: “Why have there been numerous recriminations against the directive?” is a huge slap in the face of the EU public. The answer admits: “some statistics inside the European Parliament show that MEPs have rarely or even never been subject to a similar degree of lobbying before (such as telephone calls, emails etc.).” But it then dismisses that unprecedented level of protests as follows:
“There are numerous precedents of lobbying campaigns predicting catastrophic outcomes, which have never come true.
For example, telecom companies claimed phone bills would explode as a result of caps on roaming fees; the tobacco and restaurant lobbies claimed people would stop going to restaurants and bars as a result of the smoking ban in bars and restaurants; banks said they would have to stop lending to businesses and people, due to tougher laws on how they operated and the duty-free lobby even claimed that airports would close down as a result of the end of duty-free shopping in the single market. None of this happened.”
Notice that every “counter-example” it gives involves companies complaining about laws that benefit the public. But that is not the case with the wave of protests against the Copyright Directive, which come from the public, and are directed at the selfish demands of the copyright industry. The Q & A tries to make a false equivalence between the self-interested whining of lazy industries, and the attempt by worried tech experts and millions of concerned citizens to preserve the amazing power and freedom of the open Internet.
This, ultimately, is why the Copyright Directive is so pernicious – because it ignores completely the rights of Internet users. The fact that the new Q & A is unable to answer any of the serious criticisms of the legislation without resorting to shabby casuistry and word games is a confirmation that this is not just dangerous law-making, but deeply dishonest too. If Article 13 is passed, it will undermine the Internet in the EU, turn the region into a digital backwater, and taint the entire democratic system as a result of the EU’s repeated refusal to listen to the citizens it supposedly serves.