DE policy/Copyright-Forderungen in Zeiten von Corona - Replik auf Helga Trüpel

From Meta, a Wikimedia project coordination wiki

Copyright claims in times of Corona: Wikimedia Germany replies to Helga Trüpel[edit]

Over the last couple of weeks, a variety of stakeholders have published contributions in traditional and social media on the subject of copyright in times of coronavirus/COVID-19. A number of artists, publishers, and others close to copyright-based industries have called for measures to improve the situation of individuals and companies suffering from the crisis. Proposals have included an early implementation of the Digital Single Market Directive and, worse, “eternal copyright,” as suggested by the French musician and UNESCO Goodwill Ambassador Jean-Michel Jarre. Representatives from libraries and users’ rights organizations have demanded new limitations and exceptions to copyright or a more flexible application of existing limitations and exceptions to improve access to culture and education for as long as libraries, schools, and universities remain physically closed. Both problems are real. It is also true that many artists and creators are struggling as a result of gigs being cancelled and the overall economic downturn.


The longer public social life is shutdown, the more dire will the situation become for cinemas, concert promoters, theaters, etc. We agree that the state should intervene in various ways to save jobs, offset financial losses, and mitigate uncertainty to preserve the cultural sector. At the same time, the pandemic crisis highlights the need to improve digital access to textbooks and library collections. Academics and practitioners warn that particularly lower-income families grapple with the labor market effects of the crisis and will suffer the most from the cancellation of classes and library closures. Measures to facilitate the digital use of educational materials or the provision of library services via online channels, such as reading books to children, could cushion the blow.


The debate has been polarized with some participants taking extreme positions. We take this opportunity to reply to a number of select contributions, highlighting where we think copyright could play a role in solving some of the underlying issues and identifying a middle ground for future discussions. We start by reacting to a piece by former German MEP/Greens Helga Trüpel in the conservative German newspaper Frankfurter Allgemeine Sonntagszeitung, in which Wikimedia has been explicitly namechecked (A campaign against creators: virus against copyright).


The Fifth Column of Google? “Internet Activists”

In her opinion piece, Ms. Trüpel claims that

internet activists and platforms of the Silicon Valley whose goal has been for decades to weaken or completely abolish copyright have drawn new attacking energy from the corona pandemia. (our translation)

According to Trüpel, these actors …

are using the crisis - like the Internet Archive in the United States - to argue in favor of ‘free flow of information’ and ‘free access to culture’ without taking due consideration of the interests of professional authors. (our translation)

In German: Internetaktivisten und Plattformen des Silicon Valley, deren Ziel es schon seit zwei Dekaden ist, das Urheberrecht zu schwächen oder gar ganz abzuschaffen, ziehen neue Angriffsenergie aus der Corona-Pandemie, indem sie die Krise nutzen, um – wie etwa das „Internet Archive“ in den Vereinigten Staaten – für „free flow of information“ und „free access to culture“ zu argumentieren, ohne irgendeine Güterabwägung mit den Interessen der professionellen Urheberinnen und Urheber vorzunehmen.

Here, Trüpel uses the longstanding trope that internet activists and “Silicon Valley platforms” (or Google) are in cahoots to advance an anti-copyright agenda. Trüpel insinuates that civil society activists and corporate lobbyists share the same goal of making content that has been produced by hard-working artists available at no cost – and that they are exploiting the current situation to advance this goal, which is what she herself does with this piece and its more than obvious title.

The framing of internet activists being the fifth column of Google has been in use for quite some time by rightsholders to delegitimize all demands for a modernization of copyright - even sensible ones. As will become very clear as we proceed, neither Wikimedia nor most other activists aim at abolishing copyright. On the contrary, we recognize that copyright plays an important role in facilitating the remuneration of artists and creators. Moreover, the Wikimedia communities are amongst the most pedantic watchdogs regarding compliance with copyright rules and personality/moral rights.


Neoliberal Limitless Freedom and Socialist Collectivization

Trüpel proceeds to claim that – together with internet monopolies like "Youtube” and with the Pirate party – foundations including “Wikimedia” (thus probably referring to the Wikimedia Foundation, but unclear whether referring also to the chapters, user groups, and communities of the Wikimedia Projects) and the Shuttleworth Foundation were one block, arguing for “neoliberal limitless freedom” and non-liability for platforms on the one hand and the “socialist” collectivization of content on the other:

The campaign of the internet activists (ranging from the Pirates and internet monopolies like Youtube to foundations such as Wikimedia and the Shuttleworth Foundation) continues to rest on a completely unregulated notion of freedom and thereby positions itself neo-liberal regarding platform regulation (no liability, no control) and socialist regarding the users and uploaders: Everything shall be available unregulated and instantly. (our translation)

In German: Nach wie vor fußt die Kampagne der Internetaktivisten (von den Piraten, über Internetmonopole wie Youtube und Stiftungen wie Wikimedia und die Shuttleworth Foundation) auf einem völlig unregulierten Freiheitsbegriff und gibt sich damit neoliberal in Bezug auf die Regulierung der Plattformen (keine Haftung, keine Kontrolle) und sozialistisch in Bezug auf die Nutzer und Uploader: Alles soll unreglementiert und sofort zur Verfügung stehen.

First of all and to reiterate this point, Wikimedia Deutschland has never demanded the collectivization of works. If anything, we and the overwhelming majority in our communities support a copyright system that functions as a fair remuneration mechanism for authors and creators. Our communities are authors and creators themselves, something that people seeking out that all-encompassing internet activist conspiracy tend to forget or try hard to ignore. We believe, however, in the social purpose intellectual property has like any other kind of property, a purpose that is in the law itself, in the constitution even.

While one central purpose of copyright is to put money in the pockets of authors and rights holders (and thus to incentivize creativity) copyright must – by law – also enable uses of protected material that promote the common good. This is not a controversial opinion. It is the decision of the lawmakers the world over, and forms part of the balance that justifies the temporary monopolies copyright provides. Also beyond the law, practically everybody agrees that copyright laws require limitations and exceptions to work well.

What is contested is the extent of such flexibilities (regarding their scope and whether they are remunerated through levies and taxes or generally free of cost for users). Such questions have been answered by legislation on the national level, leading to a very diverse landscape worldwide. This diversity can be a massive problem when it comes to cross-border use, and lawmakers are constantly trying to harmonise rules where possible. Here, Wikimedia Deutschland and other Wikimedia organizations have leaned towards more harmonisation rather than less, towards collective rights management rather than micro transactions, and in general towards approaches that burden the financial consequences only on those who can handle them.

Assuredly, we would like to see greater leeway for educators and heritage institutions for example but so do schools and libraries. This debate has been and continues to be highly contentious. Trüpel’s polarizing approach to put everybody who does not completely agree with her in a corner with copyright abolitionists is decidedly not fruitful. It has also been one of the major reasons for the lack of productiveness of the debate on the Digital Single Market Directive that some of the differences in positions have been completely blown out of proportion. Due to this way of framing others into extreme camps, much of the opportunities that the Digital Single Market Directive initially proved were lost, for creators and users alike.


False Equivalence

The biggest problem of Trüpel’s argumentation, however, is false equivalence. Trüpel basically says that as we oppose upload filters, we welcome terrorist and other illegal content on platforms, including anti-semitism and right-wing extremism. This claim is interesting for multiple reasons.

It culminates in the wording 'Limitless Freedom.' One cannot phrase it in a more neo-liberal way than that. This Framing is more than problematic, given the developments around right-wing radicalism, anti-semitism, terror posts and non-payment for professional work on the net. (our translation)

In German: Es gipfelt in der Formulierung „Limitless Freedom“ (grenzenlose Freiheit). Neoliberaler kann man es nicht formulieren. Dieses Framing ist mehr als bedenklich angesichts der Entwicklungen von Rechtsradikalismus, Antisemitismus, Terrorposts und Nichtbezahlung professioneller Arbeit im Netz.

To even put “non-payment for professional work” on one level with political extremism and racism should have triggered whoever was the editor of Trüpel’s piece to change that sentence. But apart from this outrageous faux pas in world view and style, the factual basis is missing as well:

Article 17 of the Digital Single Market Directive (and in fact the entire Directive) addresses “intellectual property” only. Apart from national obligations under secondary liability (in German: Störerhaftung) and case-specific rulings by the ECJ, there is to this day no general monitoring obligation in EU law regarding other forms of illegal content (and thus no requirement to proactively moderate content, let alone to filter the entire upload traffic of a platform). In fact, the DSM Directive explicitly states that there must not be such a general monitoring obligation. While we do support the liability exceptions for internet platforms as established by the InfoSoc Directive two decades ago – because we think that the introduction of a general monitoring obligation is likely to lead to the overblocking of content (particularly of fringe opinions and in context-sensitive cases) and would thus hurt freedom of expression – we do not think that platforms should have no duties or obligations at all. We are not at all asking for “limitless freedom”, which is why Trüpel’s claim is simply false. We in fact support proactive moderation efforts and platform responsibility. But we argue that the solution cannot and must not be automated systems prone to overblock and hurt fundamental rights on a net-wide scale.

No filtering technology of today and the foreseeable future is able to judge a person’s online expression based on its context. The companies offering filtering technology are even openly saying that they do not invest in such a “skill”. What the systems do excel in, is recognition of content. That’s where the technology has been and is evolving rapidly. The matching of new uploads against existing content covered by copyrights. The systems nowadays even correctly identify modified versions of copyrighted material. They are excellent detectors, but terrible judges. And judgment is needed, in every single case, for every single upload. That’s what fundamental rights are all about, and that is in every single case dependent on the context in which the expression is made. Filtering systems are extremely bad regarding context. It’s not what they are made for and not what their makers case about at all.

And even if that were different, even if algorithms would be directed at micro judging the context of online expressions and would actually be able to deliver meaningful results, could anyone in their right mind want such technology deployed everywhere on the net? We don’t think so, because that would mean the de-facto abandonment of human oversight in the online world, the potential for abuse of the automated judgment system’s levers being so enormous that democracy itself would henceforth depend on the flick of a switch. We don’t want that – and neither do the proponents of better copyright enforcement through platform liability a la Art. 17, yet that is what their otherwise partly legitimate intentions could bring about.

Instead of imperfect filters we need incentives, rules and maybe even obligations for more human oversight and a better public infrastructure to support users who have been targeted by hate speech or whose contributions have been blocked with no good reason, and yes, also for those affected by copyright infringements.

We also struggle to understand what Trüpel means by the accusatory claim of neoliberalism at work. In the German political discourses, neoliberal is a framing almost consensually regarded as negative, pointing at detrimental effects of market radicalism and never ending privatisation of gains and collectivisation of losses, associated especially to economic policy since the 1980s. Maybe Trüpel tries to tap into the general distaste for that kind of neoliberal thinking. She however blanks out what is really at stake in the Art. 17 debate, apart from the market dimension that is in the DSM Directives name: Fundamental rights and freedom of expression in particular are basic principles of liberal democracy. To discredit the fight for freedom of expression as a neoliberal campaign is not only wrong, it’s dangerous.


Defense of Monopolies

Finally, Trüpel argues that this would uphold existing monopoly and hamper competition.

To not take note of the negative developments, borders on political autism. This ignorance ennobles itself by an unprecedented argumentative chain supposedly unclouded successes. Pure cultural socialism is promised. The monopolisation of digital platforms is not made a subject of discussion as a problem. (our translation)

In German: Die negativen Entwicklungen nicht zur Kenntnis zu nehmen grenzt an politischen Autismus. Sich selbst adelt diese Ignoranz durch eine beispiellose Argumentationskette angeblicher ungetrübter Erfolge. Der reine Kultursozialismus wird versprochen. Die Monopolisierung der digitalen Plattformen wird nicht als Problem thematisiert.

Ms. Trüpel should at least try to check such claims before publishing them in one of Germany’s leading newspapers (if she doesn’t like to use Google Search, other search engines also make it easy to verify what we actually put forward in the debate). Our arguments heavily rely on the observation that market dominance of few large platforms is indeed a massive problem. For example, in an extensive public reply to Trüpel’s MEP colleague Axel Voss (EPP) we specifically call for finding solutions to this problem in the field where they are rooted: Competition law and antitrust regulation. We support competition and would love to see measures that facilitate market entry, such as requirements for sharing non-personal data or interoperability. Furthermore we do not see how filters would foster competition (and Trüpel also fails to explain this). To the contrary, as could be witnessed time after time, added obligations targeting all platforms strengthen the big ones in relation to the rest, because they simply have the better means for compliance.

We also agree that platforms that benefit from copyrighted content drawing traffic to their systems should pay more of their indirect revenue to creators. But the Art. 17 approach is too harmful for fundamental rights to be the solution. In order to make the current copyright rules better enforceable, one must consider dedicated platform APIs for rightsholders, instant delisting obligations with added notice-and-notice mechanisms, and similar approaches that have been put forward by us and others. Instead, Art. 17 as it was passed turns the current liability system on its head, with potentially devastating side-effects for freedom of expression, and strengthens the already dominant players in relation to the smaller ones.

Trüpel doesn’t mention this, for her it’s simply a big campaign against creators. Maybe she doesn’t want to see the actual setup, which would mean that she’s projecting her own denial onto everyone who doesn’t share her opinion.


Grounds for Compromise

Trüpel in her piece then tries a kind of closure by asserting that the DSM Directive in effect already is a good compromise, one that sufficiently addresses the concerns of its critics, such as the ones mentioned above in this text. Whether she herself believes that, we cannot know, but it is highly doubtful whether that assertion is correct.

It is therefore right to not let algorithms by themselves decide on what can be uploaded or will be deleted. Human judgment is indispensable for this. People doing this important work must be paid by the platforms. This approach, to combine algorithms (vulgo: 'upload filters') and human judgment is part of the European copyright reform. Also, the European copyright reform is designed to regulate monopolies and to not put the same level of responsibility on small platforms or to even exclude them entirely, like online fora and for example Wikipedia. (our translation)

In German: Deswegen ist es richtig, dass Algorithmen nicht alleine darüber entscheiden dürfen, was hochgeladen oder gelöscht wird, dazu ist menschliche Urteilskraft unerlässlich. Menschen, die diese wichtige Arbeit leisten, müssen von den Plattformen bezahlt werden. Dieser Lösungsansatz, Algorithmen (vulgär: 'Uploadfilter') und menschliche Urteilskraft zu verbinden, ist Teil der europäischen Urheberrechtsreform. Ebenso ist die europäische Urheberrechtsreform darauf ausgelegt, die Monopole zu regulieren und kleine Plattformen nicht im selben Maße in die Verantwortung zu nehmen oder sogar ganz auszuschließen, wie Foren und zum Beispiel Wikipedia.

So, in the final section, Trüpel backpaddels from her attack on Wikimedia and the Shuttleworth Foundation, arguing that algorithms should be combined with human oversight. Whether the DSM Directive actually does command that, as Trüpel says, is highly debatable.

It’s not even that Trüpel cheats a little when she mentions Wikipedia as an example for small platforms being excluded. “Online encyclopedias” are excluded (through the definition of Online Content Sharing Providers in Art. 2) in general, i. e. because of their function and role, not because of size. Wikipedia as the prime examples for such encyclopedias is amongst the ten largest websites in the world, surely not a “small platform”.

Trüpels point on small platforms leads to section 6 of Art. 17, that way hailed as a protection for startups. As an exception for small platforms it is entirely dysfunctional, as it contains a hard deadline of 3 years – after which any platform, no matter how small, falls under the full obligations of Art. 17. To sell this as a great remedy for the monopoly-strengthening effects of Art. 17, as Trüpel does indirectly and others have done more explicitly, borders on satire.

Much the same applies to the purported “human oversight”. That is present in the Directive as a mere half sentence in section 9 of Art. 17, demanding that “decisions to disable access to or remove uploaded content shall be subject to human review”. The main problem with this very unspecific rule is, that it is part of the platforms’ requirement to deploy “effective and expeditious complaint and redress mechanisms”. To spell it out, the logic here is this:

Systems of whatever nature – due to economic reasons those can only be automatic ones aka algorithms – by themselves decide on what can be uploaded or will be deleted (to use Trüpel's own words) and later, retrospectively, there is some kind of process to file a complaint, and only then a human comes in and checks the machine’s decision.

Now imagine this in the context of a heated debate on a political topic: Certain interventions don’t appear in the debate’s stream of arguments on the platform due to automatic filtering, the exact reasons unknown to the ones posting them. After retrying several times to rule out connection problems etc., the person whose interventions were blocked needs to resort to some sort of complaint button, use it, wait, and probably explain why she thinks her post was blocked in violation of the proper standards regarding freedom of speech. Then, hopefully after only minutes, more realistically hours or even days, some decision is made and the post either appears or not. And now imagine this for every single blocking decision separately, and by now even the last person will have grasped why this approach is a massive impediment to freedom of expression. Outsourcing the guarantee of fundamental rights to the complaints department for later redress is a big slap in the face of everything the European Union prides itself to foster at home and abroad: Human rights, democracy, good governance, the proportionality principle, and in the end reason itself.

Trüpel’s joining in this, as a kind of self-proclaimed spearhead of everything that is good and just around Art. 17, is especially concerning given her background and roots in the German Green party. She, if anyone, should know about the value of fundamental rights, especially when in comes to political debate.

To also mention our position on algorithmic systems: We agree that algorithms can play a role particularly in flagging problematic content. They are in fact already used that way within Wikipedia, where algorithms and deep learning are used to detect vandalism. They are, after all as mentioned above, excellent detectors. The decisions to block or delete content, however, are in every single case taken by humans, in Wikipedia’s case learned volunteer moderators. These two ingredients are crucial: Human decision, each case separately. In particular, we reject any kind of staydown regime, that would not allow the re-upload of content that has once been blocked or deleted. Again, context is important and filters, the terrible judges they are, fail to recognize context.

Art. 17 actually makes a hollow attempt to take into account that user-generated content is often covered by parody or pastiche exceptions in copyright. Section 7 states:

Member States shall ensure that users in each Member State are able to rely on any of the following existing exceptions or limitations when uploading and making available content generated by users on online content-sharing services: (a) quotation, criticism, review; (b) use for the purpose of caricature, parody or pastiche.

Yet there is zero indication here as to how this is to be achieved beyond the “complaint and redress mechanisms” mentioned above, them being no effective remedy, as explained above. So, in Art. 17 we find no less than 4 sections that describe in detail what platforms must do to improve the position of rightsholders, against 1 that states things obvious in any democracy, such as that criticism must be possible, but gives no detail whatsoever how to achieve this. That is a textbook example of bad lawmaking.

The second such instance within the very same Art. 17 is its section 8. After the previous sections establish an obligation that can only be complied with by scanning all incoming uploads against rightsholder claims, section 8 boldly states:

The application of this Article shall not lead to any general monitoring obligation.

This contradictory bogus is what gives forced compromise in general and Art. 17 in particularly its bad name. Still, Ms. Trüpel would probably cite that section 8 as a means to remedy potential concerns about mass surveillance in the name of copyright. If that section will ever have any effect, it will be through a ruling of the European Court of Justice, and will then bring down the entire Art. 17. This is something Trüpel wouldn’t want, of course, and it is also very unlikely to happen, leaving section 8 a dead rule.

Nota bene: Much of what we lay out here regarding the DSM Directive and its Art. 17 also applies to the other highly controversial EU instrument of moderating online content, the Terrorist Content Online Regulation (TERREG). It follows the good intention to push back on terrorist propaganda flowing freely on the internet, but the side effects are again grave. For example a journalist’s report on terrorist activities is also likely to be blocked by a filter implemented to comply with TERREG, even the report is covered by press freedom. Filters do not recognize context, only humans do and only they should try. We have made our position very clear in the debate on the DSM and on TERREG and will continue to do so in the upcoming years when yet another specimen of filter legislation will be discussed, the already infamous Digital Services Act.


How Could Copyright Solve the Problems of Creators and Users?

There are indeed ways in which copyright could be reformed to enable a better remuneration of authors and creators and facilitate access to knowledge. The solution would be however different from the already passed Art. 17, and much different from what some rightsholders have been suggesting in recent weeks, including calls to again extend copyright terms of protection. It seems that contributions, such as Trüpel’s are a smokescreen to divert from authors’ actual problems and the calls for access to educational materials by libraries, schools, and universities. Copyright could be part of the solution for authors and creators but would have to be reformed to improve remuneration. Specifically, copyright could put artists and creators in a more powerful position to negotiate better contract terms, if they would enjoy conversion rights or termination of transfers rights all around the world.

In marked contrast to this, copyright is still not fully adapted to the digital age. Authors struggle to make a living as public readings are cancelled. Schools and universities face difficulties in putting their material online without incurring significant additional cost. Considering that these institutions face major investments in their digital infrastructure already, they should be enabled to use educational and scientific material particularly where it has been publicly funded. The same applies to libraries and museums. Culture needs to be available even if the physical institution is closed. However, these institutions cannot provide some of their most basic services due to legal uncertainty or additional licensing fees.

So yes, copyright has a role to play in making the future better – but in a very different play compared to what Ms. Trüpel and others put on stage.