Background

According to Article 14 of the CDSM Directive, any material resulting from an act of reproduction of a public domain work cannot be subject to copyright or related rights-protection unless it is original in the sense that it is the author’s own intellectual creation. All EU member states are required to implement Article 14 and amend their national legislation accordingly (see Deborah De Angelis’s blog post).

Italy transposed Article 14 of the CDSM Directive into domestic law by passing Article 32-quarter of the Italian Copyright Law n. 633/1941 in November 2023 (Legislative Decree no. 177). The article includes a problematic addition: “The provisions on the reproduction of cultural heritage contained in Legislative Decree No. 42 of 22 January 2004 [CCHL] remain unaffected.” The reference is to Article 108 of the CCHL, which requires the payment of a concession fee for the reproduction of digital images of state-owned cultural heritage in the public domain if the reproduction is for-profit. The provisions of the CCHL beg the question of the protection of the public domain by other fields of law outside of copyright and related rights, which are not explicitly mentioned in Article 14 of the CDSM Directive (see Mirco Modolo’s article on the subject). Only the European Court of Justice (ECJ) can provide guidance on this matter. However, no Italian court has referred to the ECJ yet to clarify the relationship between Article 14 and the CCHL.

In recent years, the CCHL has been used by Italian cultural heritage institutions to initiate a number of lawsuits against commercial uses of works by Italian artists, which are clearly in the Public Domain.

Recent case law on the unauthorised reproduction of Italian cultural heritage in the public domain

1) Ministry of Culture v. Studi d’Arte Cave di Michelangelo:

In 2018, a famous Italian luxury fashion brand posted a video on the internet in which a physical copy (a clone of the statue) of the David—created by Studi d’Arte Cave Michelangelo S.r.l. (Cave)—could be seen wearing a tailor-made high-quality outfit of the brand. The Ministry of Culture sought to have the Court of first instance of Florence issue an interim measure to prevent further use of the image of Michelangelo’s David for commercial purposes. The Court dismissed the petition on the grounds of lack of urgency, as both defendants had removed the contested material from their websites.

However, the Ministry of Culture found out that Cave continued using the image of the David on another website, studidarte.it, still for commercial purposes, and filed a new urgent petition against Cave in 2021. The petition was dismissed and then appealed against it. On 11 April 2022, the Court ordered (see Simone Aliprandi and Carlo Piana’s comment) Cave to stop using the images of the David for commercial purposes, to remove all images of the statue from Cave’s websites, to pay the Gallerie dell’Accademia di Firenze a fine of € 500 for each day of delay in the execution of the preventive order, and it was further ruled that summaries of the order would be published at the expense of the defendants in two national daily newspapers, in two local daily newspapers and on Cave’s Instagram profile and YouTube channel.

The Court concluded that the mere ex-post payment of compensation is insufficient for the legitimate reproduction of a cultural asset. For the use of the image to be lawful, consent is required, following a discretionary assessment of the requested use (and its possible configuration) concerning the asset’s cultural purpose and historical-artistic character. The nature of a cultural asset inherently requires the protection of its image through an evaluation of compatibility reserved for the Public Administration. This evaluation encompasses the right to reproduce the asset and the safeguarding of the asset’s consideration by fellow citizens – its identity as a collective memory of the national community and the territory. Therefore, according to the Court of Florence, this notion should constitute a comprehensive right to the cultural asset’s image (right of publicity).

2) Galleria dell’Accademia Firenze v. GQ:

In 2020, the Gallerie dell’Accademia di Firenze and the Ministry of Culture sued the publishing house GQ (Condé Nast) for the unauthorised use of the image of Michelangelo’s David on the July/August issue’s cover of GQ Italia (see Justus Dreyling’s post on this blog and this interview with Deborah De Angelis). The plaintiff asked the Court to enjoin the use of the image of Michelangelo’s David. The Court promptly issued an order banning the use of the image on the cover of the magazine and prohibited any further digital use of the image.

Eventually, on 15 May 2023, the same Court ruled again in favour of the Gallerie and the Ministry of Culture and condemned the publisher GQ to pay the Galleria dell’Accademia di Firenze two separate amounts: € 20,000 as a concession fee and an additional € 30,000 for the way in which David’s image was distorted for the magazine. In addition, the judge acknowledged the right to the image (which is granted by Article 10 of the Italian Civil Code to physical persons and legal entities), with specific reference to cultural heritage, considering the legal basis for this right is found in Articles 107 and 108 of Legislative Decree No. 42/2004, which directly implement Article 9 of the Constitution(See Eleonora Rosati’s comment on IPKat).

3) Gallerie dell’Accademia di Venezia v. Ravensburger:

On  24 October 2022 (see Deborah De Angelis and Brigitte Vézina’s comment on this blog and Giuilia Dore’s contribution on the Kluwer Copyright blog), the Court of first instance of Venice decided on the lawsuit brought by the Gallerie dell’Accademia di Venezia, a public museum under the Italian Ministry of Culture, against the German toy-making companies Ravensburger AG and Ravensburger Verlag GmbH as well as their Italian branch represented by Ravensburger S.r.l. for the unauthorised use of the images of Leonardo da Vinci’s Vitruvian Man on a series of puzzles. The order enjoined the German company to stop using the image of the Vitruvian Man for commercial purposes, to pay to the Gallerie dell’Accademia di Venezia a fine of € 1,500 for each day of delay in the execution of the preventive order, and it was further decided that summaries of the order would be published at the expense of the defendants in two national daily newspapers and in two local daily newspapers.

4) Uffizi v. Gaultier:

In October 2022, the Uffizi Galleries announced their intention to sue the French fashion house Jean Paul Gaultier for damages that could exceed € 100,000 after the company’s (allegedly) unauthorised use of images of Botticelli’s Renaissance masterpiece The Birth of Venus to adorn a range of clothing products, including T-shirts, leggings, and tops, for the brand’s new line Le Musée. There has not been any news on this controversy, and there is no evidence that a lawsuit has been served (see, Justus Dreyling, Brigitte Vézina, and Teresa Nobre’s post on this blog).

It is clear that a protectionist trend is emerging in line with the approach of the government (for a critical approach, see Roberto Caso’s comment on the Kluwer Copyright blog).

The Ministerial Decree on the minimum tariff for the reproduction of the digital images of the state cultural heritage

Prior to 11 April 2023, the interpretation and practice of art. 108 of CCHL granted cultural heritage institutions discretion to decide whether to adopt an Open Access policy, enabling the use of the digital images of cultural heritage through the use of a Creative Commons licence or Public Domain tools, to promote fair access to and sharing of Italian culture, while supporting the role of cultural heritage institutions in sustainable economic and social development.

In April 2023, however, the Italian Ministry of Culture  introduced minimum fees for commercial reproductions of state-owned cultural heritage, including for works in the Public Domain that all state-owned public museums will have to apply (Decree no. 161 also known as Guidelines). The Decree will have a detrimental effect on the promotion and dissemination of Italian cultural heritage globally, impeding knowledge sharing (see reporting on huffingtonpost.it and repubblica.it). The new guidelines represent a significant setback as they contradict the fundamental principles of public enjoyment and enhancement of cultural heritage enshrined in the Italian Cultural Heritage Code.

The situation in Italy escalated on 14 June 2023, when Senator Marcheschi (Fratelli d’Italia) proposed to punish with a fine between € 20,000 to 60,000 the unlicensed use of cultural objects in the public domain in violation of Articles 107(1) and 108(1)-(3) of the CCHL (as an amendment to the proposal of law on Article 518-duodecies of the Italian Criminal Code, on the destruction, dispersal, deterioration, defacement and illegal use of the physical cultural heritage and landscape). At the session of June 21st, 2023, the amendment was withdrawn, having received a negative opinion because of the costs it would entail to activate the sanctions, but it turned into an agenda for the Government “to consider the advisability of providing for the imposition of an administrative fine of 20,000 to 60,000 euros against anyone who, in violation of Articles 107, paragraph 1, and 108 paragraphs 1, 2 and 3 of Legislative Decree No. 42 of January 22, 2004, reproduces a cultural heritage or markets its reproduction in the absence of or in contravention of the order of the authority in charge of the property.”

A solution for no problem

The flow of income generated from the licensing of images of cultural objects remains more or less unchanged for major museums. Peripheral and smaller museums, by contrast, lament an increase in bureaucracy not supported by the hiring of new staff. The reproduction and dissemination of images of cultural heritage in smaller museums, even for commercial purposes, contributed to the diffusion of culture and the promotion of national heritage.

The discretion that was left to individual museums when licensing images of the objects under their custody, allowed the museum staff to consider different factors: internal costs, whether they already had good quality images not covered by copyright, the promotion of the museum, the valorization of the cultural object, etc. The mandatory application of the minimum tariff stated by the guidelines makes it impossible, e.g. for state-owned museum staff, to permit the free use of images according to the open access principle. In the past, museums and other state-owned cultural institutions could allow the free use also for commercial purposes (as permitted by CC licences and tools compatible with open access) without asking for a concession fee. The discretion previously held by cultural institutions to decide whether to authorise the free use of cultural heritage images has been eliminated.

Instead of creating positions to help under-staffed museums or promoting the preservation of the Italian cultural heritage, the Ministry of Culture has enacted unnecessary restrictions. The imposition of high fixed fees for the for-profit use of images of cultural objects in the Public Domain may result in limiting the exploitation only to privileged classes of individuals, while preventing local communities or “communities of origin” with a limited budget from participating in it. Moreover, the community has not been involved in the decision-making process and has not had a chance to participate in the debate. This scenario is further complicated by the possibility for the licensing administration to deny for-profit uses of images by invoking the “decorum” exception, namely judging the declared use of the image of the artwork as inappropriate. “Decorum” is a very broad, subjective, and undefined concept that contributes to generating uncertainty in this field (see Daniele Mancorda’s contribution).

What is more, creativity is at risk! Today, when using professional photographic reproductions of works in the Public Domain, which are subject to both copyright and the Guidelines, users will face the so-called “tragedy of anticommons” (a term coined by Michael Heller). These works will be used less due to the existence of multiple layers of protection, which are hard to navigate.

The post Tales of public domain protection in Italy appeared first on COMMUNIA Association.

We discuss the considerations of the use of copyright-protected works and other protected subject matter as training data for generative AI models, and provide two recommendations for lawmakers. Here, we leave aside questions relating to the output of AI models (e.g. whether the output of generative AI models is copyrightable and in how far such output can be infringing exclusive rights), which we will address in another, yet to be published paper.

This paper is without prejudice to the position of COMMUNIA or individual COMMUNIA members regarding this discussion in other jurisdictions.

The post Using Copyrighted Works for Teaching the Machine – New Policy Paper appeared first on COMMUNIA Association.

This blog post analyses the implementation of the copyright exceptions for Text and Data Mining, which is defined in the Italian law as any automated technique designed to analyse large amounts of text, sound, images, data or metadata in digital format to generate information, including patterns, trends, and correlations (Art. 70 ter (2) LdA). As we will see in more detail below, the Italian lawmaker decided to introduce some novelties when implementing Art. 3, while following more closely the text of the Directive when implementing Art. 4.

Text and data mining for scientific purposes

Art. 70 ter (1) LdA permits reproductions made by research organisations and cultural heritage institutions for scientific research purposes, for purposes of extraction of text and data from works or other materials available on networks or databases to which they lawfully have access to, as well as communication to the public of the results of research when expressed in new original works. It seems that the Italian lawmaker committed a lexical fallacy by adding that the purpose of the reproduction is to extract text and data from works and other materials available in networks or databases. This, as well as limiting the type of protected materials to those present on networks or databases, appear to be legislative inventions incompatible with the directive.

Notably, the new Italian exception also allows the communication to the public of the research outcome when such outcomes are expressed through new original works. In other words, the communication of protected materials resulting from computational research processes is permitted, provided that such results are included in an original publication, data collection or other original work.

The right of communication to the public was not contemplated in the original government draft; it was introduced in the last version of the article to accommodate the comments of the Joint Committees of the Senate and the Joint Committees of the Chamber, both highlighting the need to specify that the right of communication to the public concerns only the results of research, where expressed in new original works.

The beneficiaries of the TDM exception for scientific purposes are research organisations and cultural heritage institutions. Research organisations essentially reflect the definition offered by the directive. These are universities, including their libraries, research institutes or any other entity whose primary objective is to conduct scientific research activities or to conduct educational activities that include scientific research, which alternatively:

• operate on a non-profit basis or whose bylaws provide for the reinvestment of profits in scientific research activities, including in the form of public-private partnerships;
• or pursue a public interest purpose recognised by a European Union member state (Art. 70 ter, (4)LdA).

If commercial enterprises exercise a decisive influence, such as allowing access on a preferential basis to the results generated by scientific research activities (Art. 70 ter, (5) LdA), an organisation will not be considered a research organisation under this law. Cultural heritage institutions are defined broadly, including libraries, museums, and archives, as long as they are open to the public or accessible to the public, also those belonging to educational institutions, research organisations and public broadcasting bodies, as well as the institutes for the protection of film and sound heritage and the public broadcasting bodies (Art. 70 ter, (3)LdA).

The copies of works or other subject matter created need to be stored with an adequate level of protection and can be kept and used only for scientific research purposes, including the verification of research results (Art. 70 ter, (6)). Rightsholders are authorised to apply measures to ensure the safety and integrity of the networks and databases where the works or other subject materials are hosted (Art. 70 ter, (7) LdA). Such measures shall not go beyond what is necessary to achieve and can also be defined based on agreements between right holder’s associations, cultural heritage institutes and research organisations (Art. 70 ter (8) LdA). Such negotiations have not taken place so far.

The Italian lawmaker did not expressly contemplate any specific and fast procedure for cases where technical protection measures prevent a beneficiary from carrying out the permitted acts under both TDM exceptions. However, the law now recognises to the beneficiaries the right to extract a copy of the material protected by technological  measures in certain cases. Under Art. 70-sexies, LdA, beneficiaries of the TDM exception for scientific purposes (as well as the beneficiaries of the exception for digital and cross-border teaching activities exception) shall have the right to extract a copy of the protected material, when technological measures are applied based on agreements or on administrative procedures or judicial decisions. In order to benefit from this right, the person shall have lawful possession of copies of the protected material (or have had legal access to them), shall respect the conditions and the purposes provided for in the exception, and such extraction shall not conflict with the normal exploitation of the work or the other materials or cause an unjustified prejudice to the rights holders.

While the Italian lawmaker should be praised for introducing such a right, the Italian implementation still missed an opportunity to update the controversial and rarely applied negotiation and mediation procedures set forth in Art. 71 quinquies, (2) LdA (which transposed Art. 6(4) of the Infosoc Directive) to the actual needs of all beneficiaries of copyright exceptions. Furthermore, it should be noted that the law continues to reserve the application of the three-step test only to  the exceptions and limitations related to the use of protected materials available through on-demand services only, including under the new copyright exceptions, contrary to what is mandated by Art. 7(2) of the CDSM Directive.

Regarding contractual override protection, as stated by Art. 7(1) of the CDMS Directive, conflicting agreements with the TDM exception for scientific purposes are void.

Text and data mining for other purposes

The implementation of the exception for text and data mining for purposes beyond scientific research by any individual or organisation follows for the most part the text of Art. 4 of the CDSM Directive. Reproductions and extractions are allowed from works or other subject-matters available online or databases to which users have lawful access for TDM. The exception is subject to the possibility of rightsholders reserving their rights, but there is no clarification on how this reservation should be made. In this regard, the Italian law doesn’t mention the need​​ to express such reserves appropriately, such as through machine-readable standards when contents are made publicly available online.

No additional conditions or precisions are established on the retention and storing of copies of materials made to conduct text and data mining, nor on the measures that rights holders may take to ensure the safety and integrity of the networks and databases where the materials mined are hosted. Reproductions and extractions may be retained only for the time necessary for text and data mining as stated in  Art. 4 of the CDSM Directive.

Conclusion

In conclusion, the Italian legislator should be praised for going beyond what is required by art. 3 of the CDSM Directive. The Italian implementation allows the application of the TDM exception for scientific purposes to the reproduction for the extraction of texts and data or other subject matters (to which the beneficiaries have lawful access) and to the communication to the public of the research outcome when it is expressed through new original work. Therefore, in this respect, the Italian legislator exercised the option allowed by Art. 25 of the CDSM Directive to adopt a broader provision, compatible with the exceptions and limitations provided for in the InfoSoc and Database Directives.

For TDM for other purposes, the lawmaker reproduced the text of Art. 4 of the directive, not adding any extra elements to regulate, as written before, the exercise of the faculty to opt out by rights holders in a suitable and machine-readable by the computers. The Italian implementation also missed an opportunity to introduce provisions to allow the reproduction of material protected by technical measures that prevent beneficiaries from carrying out the permitted acts under the both TDM  exceptions. In sum, a total organic revision of the Italian copyright law would be desirable, which would allow for a better understanding and knowledge by not only the interpreter of the law, but also by rightsholders and users.

The post The Italian Implementation of the New EU Text and Data Mining Exceptions appeared first on COMMUNIA Association.

This is Part 2 of a two-part contribution highlighting the report’s most significant findings. While Part 1 (published on Kluwer Copyright Blog) focused on Article 15 CDSMD, Part 2 considers Article 17 CDSMD. The report’s findings as regards Article 17 CDSMD were presented at the first session (“Fragmentation or Harmonisation? The impact of the Judgement on National Implementations”) of the Filtered Futures conference co-organised by COMMUNIA and Gesellschaft für Freiheitsrechte on 19 September 2022. 

The post is published under a Creative Commons Attribution 4.0 International licence (CC BY 4.0).

Subject matter and right-holders

As opposed to Article 15 CDSMD, Article 17 CDSMD does not introduce a new related right to EU copyright law. Instead, it expands the protections already afforded by copyright and related rights law. To this end, Article 17(1) links to Article 3(1) and (2) of the Information Society Directive (ISD). While Article 3(1) ISD covers copyright, Article 3(2) ISD lists four related rights: those of performers, phonogram producers, film producers and broadcasting organisations.

Based on the national reports, from among the 11 examined Member States, only two (Malta and the Netherlands) appear to have restricted protection to these four related rights. Denmark explicitly extends protection to producers of photographic pictures and producers of catalogues – i.e., related rights owners not mentioned in Article 3(2) ISD. Other Member States eschew a closed enumeration in favour of a general reference to related rights – implying that all related rights recognised in their national law are covered. To the extent that such rights fall outside of the EU acquis such gold-plating is arguably unproblematic. However, where harmonised related rights – such as the new PPR – are affected, implementations start slipping out of compliance.

France presents an added twist, given the absence in the French transposition of Article 15 CDSMD of an exception for private or non-commercial uses by individual users, as the Directive requires (see Part 1 on Kluwer Copyright Blog). The logical conclusion is that in France OCSSPs may be liable for non-commercial public sharing of press publications by end-users on their platforms – contrary to Article 15’s intended focus on own use by providers such as news aggregators and media monitoring services (see Recital 54 CDSMD).

Exclusive rights

Further issues arise with regard to the implicated exclusive rights. The Irish implementation of Article 17 is particularly perplexing in this regard: as the Irish national expert (Giuseppe Mazziotti) explains, while the transposition’s language is ambiguous and hard to decipher, the Irish implementation appears to require that OCSSPs also obtain authorisation from the owners not only of the right of communication to the public and of the making available right (as the CDSMD requires), but also of the reproduction right. Adding to the confusion, the immunity of Article 17(4) is transposed correctly, so that it is restricted to acts of communication to the public.

A number of possible interpretations exist. One option would be that absent authorisation from the holders of the reproduction right, OCSSPs are liable for communication to the public unless the immunity applies – though this would make little sense. An equally disconcerting possibility would be that there is no way out of a liability for unauthorised acts of reproduction that has no foundation in the Directive. Of course, it is also possible that the reference to the reproduction right is just a transposition glitch that should be ignored, in line with the Marleasing principle. This interpretation would be the most likely – were it not for the fact that it is very hard to say that OCSSPs (whose very definition in Article 2(6) CDSMD describes them as “storing” content) perform acts of communication to the public without also (at least temporarily) copying the relevant works: the absence of any reference to the reproduction right in Article 17 makes little sense. The omission has been belatedly recognised by the European Commission in its Guidance on Article 17, according to which the

acts of communication to the public and making content available in Article 17(1) should be understood as also covering reproductions necessary to carry out these acts.

The Guidance notes that Member States should not provide for an obligation on OCSSPs to obtain an authorisation for reproductions carried out in the context of Article 17. However, the Guidance is non-binding and, moreover – from a certain perspective – just confirms that acts of reproduction are inherent to acts of communication or making available to the public.

Authorized uploads by end-users

Interestingly, France has ignored the suggestion in Recital 69 of the Directive that authorisations granted to users to upload protected subject matter extend to OCSSPs. As Valerie-Laure Benabou, the French national expert, points out, the key issue here is whether OCSSPs are participating in a single act of communication to the public or making available to the public performed by the end-user or whether there are two independent acts, one performed by the user and the other by the intermediary.

Given that the Directive establishes primary and not accessory liability for OCSSPs, there is a strong argument that there are two acts of communication to the public. Of course, this seems counter-factual, as clearly only one material act occurs – the upload by the user onto the OCSSP’s platform. But that is a problem embedded in the entire solution adopted by Article 17, which holds OCSSPs liable for those uploads as primary infringers. Moreover, while recitals do have interpretative value, they do not have binding legal force and therefore cannot be used to counteract the operative part of a directive. As a result, clarity on this issue can only come with CJEU intervention.

Interaction with other types of liability

Another headscratcher is provided by the Spanish implementation. This contains a provision according to which, as has already been reported on Kluwer Copyright Blog by the Spanish national expert, Miquel Peguera, even when OCSSPs abide by the conditions of Article 17(4), right holders will be still able to rely on other causes of action for compensation, such as unjust enrichment.

How to assess this is not obvious. One view is that the provision digs a hole under the entire concept of immunity. The result, therefore, is an interference with the “occupied field” of the Directive in a way that erodes its useful purpose: the same claimant would be granted the same protection against the same OCSSP for the same behaviour for which the directive provides immunity.

As persuasive as this interpretation is, there is also a different perspective from which the provision is unproblematic. It is important to consider the difference between Article 17 and the hosting safe harbour Article 6 of the Digital Services Act (DSA) (previously Article 14 of the E-Commerce Directive): the wording of the latter clearly indicates its purpose to ensure that the relevant providers are, in a harmonised way, not held liable across the EU under certain conditions. By contrast, Article 17 arguably has the opposite objective: to ensure that OCSSPs are, in a harmonised way, held liable across the EU under certain conditions. If that is the case, then it could be said that the Spanish provision avoids interference with Article 17’s useful purpose.

Of course, even if that is so, it is highly unlikely that any OCSSP that meets the conditions of Article 17(4) would fail to meet the conditions set by any other national rule of law. The hosting safe harbour will provide additional refuge. While the issue is therefore likely a storm in a teacup, the theoretical question is an interesting one: does Article 17 exhaust the liability of OCSSPs for infringing copyright/related rights over uploaded content – does it provide total harmonisation of OCSSP liability for copyright-infringing content uploaded by their users or only total harmonisation of OCSSP liability for communication to the public?

Filtering and general monitoring obligations

Perhaps the most interesting question that arose from the study concerns the different national approaches to the implementation of Article 17(4) CDSMD. The issue was brought before the CJEU by Poland in an action for the annulment alleging that Article 17(4) makes it necessary for OCSSPs to adopt filtering technology to the detriment of end-users’ freedom of expression. In its decision, the CJEU accepted that the prior review obligations imposed by Article 17(4) on OCSSPs do require the use of filtering tools, but concluded that the resultant limitation on freedom of expression is acceptable because of the safeguards embedded in sub-paragraphs 7, 8 and 9. The Court refrained, however, from a detailed explanation on how these safeguards should operate in practice.

Among the examined Member States, Germany and Austria have taken a proactive elaborative interpretation geared at establishing how these safeguards can be put into practice. This approach has been described as a “balanced” one, as opposed to the more “traditional” one taken by those countries that adopted a copy-out transposition.

The question that arises is whether this “balanced” approach is compatible with the directive. As others have suggested, there is a strong argument that – to the extent that Article 17 does not explain how its various sub-paragraphs are intended to interact or how it’s intended to navigate freedom of expression – such elaboration might be necessary. As mentioned in Part 1, to be copiable in national law, it is necessary that a directive be internally consistent and well formulated. When provisions are in need of legislative repair or clarification, literal transposition should be dismissed. It can be said that Article 17 is precisely such a provision – as the analysis above reveals, it is not a well-drafted piece of legislation: it raises many questions, its wording is complex and confusing, and its purpose and scope are ambiguous.

That said, following the decision of the CJEU in Poland it is hard to hold that those Member States that have taken a copy-out approach to transposition were wrong to do so. Instead, the answer can be found in the final paragraph of Poland. This emphasises both that Member States themselves must transpose Article 17 in such a way as to allow a fair balance and that, “the authorities and courts of the Member States” must interpret their transpositions so as to respect fundamental rights. In this way, the Court has enabled the compatibility with the Directive and the Charter of both the “balanced” and “traditional” implementation approaches: with the first the legislator takes on the task of identifying the appropriate “fair balance” itself, via the process of transposition. With the second, the details on the right balance are delegated to judicial interpretation. The first approach may be preferable from a policy perspective – but both are open to the Member States.

Conclusion

Examination of the national implementations of Articles 15 and 17 reveals a number of potential incompatibilities. While some of these can perhaps be attributed to national intransigence or misplaced inspiration despite clear EU requirements (for example, the French failure to protect private or non-commercial uses of press publications by individual users is hard to explain otherwise), very often the root cause can be found in bad drafting by the EU legislator. The contentious subject matter and intricate structures of these articles, as well as their heavy reliance on undefined terminology and the occasional misalignment between the recitals and operative texts, are not designed to facilitate smooth national implementation and homogenous interpretation and application. As the European Commission has acknowledged, “[b]etter law-making helps better application and implementation”. In the intense discussions on Articles 15 and 17 in the run-up to the adoption of the CDSMD, this principle appears to have fallen by the wayside. To address the consequences, the CJEU will no doubt have much CDSMD-focused work ahead of it. Hopefully, future judgments will be clearer than Poland. 

The post Comparative Report on the National Implementations of Article 17 of the Directive on Copyright in the Digital Single Market appeared first on COMMUNIA Association.

The changes brought by the CDSMD to the copyright paradigm cannot be underestimated. In addition to changing the liability regime for online content-sharing service providers (see Jütte), the  obligation to introduce ex ante and ex post content moderation has the potential to turn copyright into a censorship tool if not adequately balanced against the right to freedom of expression. In the long run, the implementation of article 17 CDSMD bears crucial consequences for cultural diversity, creativity and the ability to participate in political and societal discourse online.

A strong commitment to respecting the right to freedom of expression

Whilst a strong commitment to freedom of expression is commended, the balance to be struck between the right to property of copyright holders and the right to freedom of expression remains uncertain, as the EU legislator and the CJEU in C-401/19 leave this matter for EU member states to resolve. The potential serious interference with the right to freedom of expression and freedom of the arts (guaranteed by articles 11 and 13 of the EU Charter of Fundamental Rights respectively) created by inadequate controls on content moderation is undeniable, as acknowledged by the AG (at 77 and 151). The AG also confirmed that this interference is ‘attributable’ to the EU legislature and should not be left in the hands of service providers (AGO, at 83-84). On the compatibility of this limitation with the right of freedom of expression, the AG went back to the basics by first reminding that freedom of expression does not constitute an absolute right but amounts to a qualified right, meaning that limitations to the exercise of this fundamental freedom are possible as long as they meet the three-pronged test of: (1) being ‘provided by law’; (2)  respecting the ‘essence’ of that freedom (see O’Sullivan) and, (3) respecting the principle of proportionality (in accordance with art. 52(1) of the Charter which echoes the traditional test of art. 10(2) ECHR, as confirmed by the AG at 90).

This post wishes to focus on the third step of this assessment. To determine whether the limitation on freedom of expression is proportionate, the AG first recalled that over-blocking represents an inherent risk, and that current content moderation can only look for ‘matches’ rather than identify copyright infringement (see Jacques, Garstka, Hviid and Street). As such, this obligation must be accompanied by sufficient safeguards without which art. 17 would not constitute a proportional limitation on freedom of expression. Arguably, such safeguards are provided by paras 5, 7, 8 and 9 of art. 17 (AGO at 155). Here, through a combination of the addressees of art. 17, the mandatory character of some copyright exceptions such as parody, and the requirement of a complaint mechanism to protect users’ rights, the AG finds that proportionality has the potential to be met if adequately implemented in practice to achieve the goals of the provision (at 204 and 220). Although the CJEU did not embark on a sufficiently detailed analysis to determine whether content moderation constitutes an appropriate and necessary limit on freedom of expression to meet the objectives sought, the Court mostly agreed with the AG’s opinion whilst acknowledging that, in practice, the balance of fundamental rights and, consequently, the proportionality requirement, hinges heavily on national implementation (at 99).

Understanding the margin of appreciation granted to national authorities

Given that we are in the midst of national implementation (still) of this controversial provision, understanding what kind of margin of appreciation lies in the hands of national legislators is essential.

Firstly, it is quite noteworthy that the CJEU has not demonstrated that these filters could effectively meet the objectives set by art. 17. If filters are intended, even if not explicitly mentioned in the adopted text, arguably the EU legislator or CJEU should have investigated whether the technology is up to the task.

Secondly, if it is demonstrated that content moderation can meet the objectives of art 17 to –  let’s say – 95% accuracy, some categories of expressions might be more disproportionately hit than others, which has a detrimental impact on the diversity of expressions present on these platforms. This is likely to be the case for parodic expressions (understood here as including caricatures and pastiches) given that this type of expression can be adversely affected by content moderation. This is because such uses imply copying by their nature and are precarious particularly when carrying offensive messages. In addition, parody cases decided by EU member states under art 5.3(k) InfoSoc Directive are rather inconsistent in their application of the copyright exception as to the type of uses covered (questions as to how much copying or what type of humour is allowed remain the subject of many domestic discussions). Such inconsistencies can even be found within a single member state between different courts, making it extremely difficult for OCSSPs to assess the scope of the exception or for users to justify their use and save their expressions from over-blocking.

One of the possible consequences of leaving too wide of a margin of appreciation to member states is that the safeguards intended by art. 17 result in placing too much burden on users to demonstrate the legitimacy of their use, as is seen in some member states where the broadening of the interpretation of the parody exception has led courts to require defendants to demonstrate in what manner the copying of copyright-protected content was necessary (see e.g., Malka v. Klasen saga in France). This must be avoided as it also goes against the mechanics of art. 10 ECHR, as incorporated into EU law by arts 11 and 52(3) of the Charter.

It has already been argued that the way in which the ECtHR has approached the use of humour and balanced the right to freedom of expression with the right to property, as well as other fundamental rights such as the right to protect one’s reputation, can help in colouring the application of requirements of the parody exception to achieve harmonisation and the necessary predictability (Jacques). But beyond the application of the parody exception, the ECHR framework and its jurisprudence influence the margin of appreciation left to member states in their implementation of art. 17 to ensure that the new obligations contained therein remain proportional.

When faced with conflicting fundamental rights, the CJEU reminds us that a fair balance must be struck between the fundamental rights at stake. Looking at how the balance should be struck between the users’ right to freedom of expression and the property rights of right-holders, the ECtHR’s jurisprudence becomes helpful. Whilst, in the last decades, the jurisprudence has over-emphasized the margin of appreciation granted to member states, more recent decisions operate a shift in trying to recalibrate the balance so that any restriction to freedom of expression is more predictable and the subjectivity of national judges less influential. The ECtHR has had the opportunity to establish that the member states’ margin of appreciation is wider when it comes to purely commercial expressions than it is for political or artistic expressions (see Sekmadienis Ltd. v. Lithuania, para 73; markt intern Verlag GmbH and Klaus Beermann v Germany, para 33).  Furthermore, the ECtHR has begun to be stricter with Member States in emphasising that the intent and context of an expression are appropriately taken into consideration (See Patrício Monteiro Telo de Abreu v. Portugal, paras 37, 42 and 43). As held by the ECtHR, in a case regarding political caricatures, when balancing freedom of expression with the right to protect one’s reputation, the member state should not excessively focus on the right to protect one’s reputation at the expense of the exercise of freedom of expression. In an even more recent case dealing with criticism of the Bible made by a Polish popstar on the radio (Rabczewska v Poland, paras 58-59), the ECtHR noted that one must consider the normal audience of the speaker to determine whether a restriction is justified.

A shift towards a framed margin of appreciation and a right of being heard online

This shift towards a framed margin of appreciation is of relevance when considering how art. 17 CDSM Directive will operate in practice to ensure that proportionality is met and that a vibrant digital public sphere remains possible. Indeed, an overview of national cases where courts have had to strike a fair balance between the right to freedom of expression and the right to property has often tilted the balance towards the right to property with relative ease. This may be due to the legal tradition of specific member states, but a sustainable digital environment calls for greater scrutiny of this balancing exercise. It may also be due to the over-emphasis on the ability to express oneself freely in the digital environment, discarding that a strong commitment to freedom of expression also implies a right to be heard. A user should have a realistic expectation of being able to be heard in the digital realm. And yet, provisions such as art. 17 CDSMD have the potential to further curtail the ability of being heard online. Putting greater emphasis on the right to be heard could lead to ensuring that the balance between fundamental rights in the context of art. 17 is appropriately struck by giving more weight to users and focusing the burden of proof on right-holders. Where an expression has been caught through over-blocking, a greater commitment to the right to be heard could give ammunition to private actors to authorise the use until a decision is reached through an out-of-court mechanism or judicial decision. In essence, this focus on the right to be heard online would facilitate the satisfaction of the proportionality requirement and bring parties on par.

Finally, but nontrivially, the pledge for a strong right to freedom of expression in the digital environment has led the EU legislator to make the parody exception mandatory, meaning that member states do not have a choice but to introduce a parody exception for the scope of art. 17. However, some member states might not have elected to implement a parody exception under the InfoSoc Directive. This could ultimately lead to a situation where platforms that do not fall within the scope of art. 17 end up deploying such content moderation without adequate safeguards for users’ rights and leaving the users in a defenceless position unable to rely on the parody exception in that member state. Hence, national legislators have been called upon by scholars to take the opportunity of implementing the CDSM directive to introduce a broader parody exception given its roots in the right to freedom of expression and the ECtHR jurisprudence which consistently reminds us that the means of communication should not matter. It is important to remember that currently =fragmentation is still possible even amongst platforms where a lawful parody is shared if the platform is not falling within the scope of art. 17 CDSM Directive and where no conventional parody exception exist under the national copyright legal framework.

The post A Two-Tier System for Freedom of Expression: Towards a Right to be Heard? appeared first on COMMUNIA Association.

On the 16th of April, the Court of Justice (CJEU) delivered its decision in case C-401/19  The Republic of Poland v European Parliament and the European Council concerning Article 17 of the Copyright in the Digital Single Market Directive (the CDSM). The Polish challenge centred on the argument that Article 17(4)(b) and (c) require ex ante preventative monitoring of all user uploads by content sharing service providers and is incompatible with Article 11 of the European Charter of Fundamental Rights. As such, the CDSM is silent on the need for ex ante preventative monitoring. Rather, Article 17(4)(b) and (c) require such providers employ their ‘best efforts’ in accordance with ‘high industry standards of professional diligence’ to ensure the unavailability of protected works. Prior to the Polish challenge, the conclusion reached by commentators on Article 17 – in particular the academic community – was that these requirements pointed in one direction only – the adoption of ex ante filtering technology.  Naturally, this conclusion did not manifest in a vacuum. Indeed, it is well known that the concern of European policymakers to remedy the so-called ‘value gap’ underpinned Article 17 and a desire that content sharing service providers follow the likes of YouTube in deploying filtering solutions such as its ContentID system. Indeed, the final non-prescriptive wording of Article 17 reflects a compromise over concern that expressly mandating for filtering technology was a step too far in burdening content sharing service providers and their users alike.

The Polish Challenge: Throwing Down the Gauntlet to the Court of Justice

Against the backdrop of this implicit European policy embrace of algorithmic enforcement of copyright, the Polish challenge can be viewed as a throwing down of the gauntlet before the CJEU to make this implicit policy express, or annul Article 17 in part or in full. The choice facing the Court was to fill the policy gap leading to the compromise wording of Article 17 and justify upload filters under a fundamental rights framework or identify alternative preventative measures that would comply with the ‘best efforts’ criterion of Article 17(4)(b) and (c). In this respect, the challenge rested on the widely held assumption that upload filters would be incompatible, in particular, with the norms of freedom of expression – not least, because current filtering technology is incapable of making a sufficient distinction between legal and illegal uses of content relative to the exceptions and limitations under the copyright acquis. In light of this, the Polish challenge asserted that Article 17 was absent sufficient safeguards to protect internet user rights of freedom of expression. Recognising that Article 17 does mandate the use of upload filters, the CJEU rejected the Polish challenge, confirming the compatibility of Article 17 with Article 11 of the European Charter of Fundamental Rights and laying considerable emphasis on the safeguards that exist within Article 17.

These safeguards are intended to operate both ex ante and ex post. Under Article 17(7), prior to deploying a filtering solution, cooperation between content sharing providers and right holders, i.e. the provision of information from the latter to the former, shall not result in the blocking of legal content. Article 17(4)(b) directs that right holders provide relevant and necessary information, which according to the Court protects freedom of expression, as content sharing providers will not, in the absence of such information, make such content unavailable. Moreover, the Court reiterated that a system that cannot sufficiently distinguish between lawful and unlawful content, in particular content that would require a value judgement as to its illegality, would not be compatible with Article 11 of the Charter. In short, the Court narrowed the scope of Article 17 as applying only to manifestly infringing content. Article 17(9) establishes the first ex post safeguard, in the event that content is ‘erroneously or unjustifiably’ blocked, by requiring content sharing providers to put in place effective and expeditious complaint and redress mechanisms. Moreover, the Member States are to ensure there is access to out-of-court redress mechanisms and recourse to the courts. The Court was therefore satisfied that Article 17 did – contrary to the Polish claim to the opposite – contain sufficient safeguards to ensure compatibility with Article 11 of the Charter. (Many) questions however remain, not least, the question as to the ‘essence’ of freedom of expression in this context.

The Fact of Safeguards Tells Us Nothing About the ‘Essence’ of Freedom of Expression

Adopting a classic proportionality formula, the Court’s position is supported by its jurisprudence and Article 52(1) of the Charter. To this end, Article 52(1) allows for fundamental rights to be limited if justified and subject to minimum safeguards, while Promusicae  makes clear internet users’ fundamental rights can be limited under a balancing exercise to support copyright enforcement. In turn, the Court had little difficulty in establishing that the ‘essence’ of freedom of expression in the context of Article 17 was respected, in particular, by the minimum safeguards under Article 17(7)-(9). In doing so, it drew on rulings of the European Court of Human Rights that prior restraints on freedom of expression are permissible, subject to a particularly tight legal framework. However, the fact of safeguards tells us nothing about their substance, particularly their actual or likely efficacy in practice relative to protecting internet users’ right to freedom of expression. Following the ruling, the normative position for the Member States is to now answer the question of how these safeguards are to work to ensure internet users’ fundamental rights are respected. Therein lies the problem. In framing the essence of the right as unaffected by prior restraints and insulated by minimum safeguards, the Court fails to illuminate the degree to which such restraints may impact the right or indeed to what degree the minimum safeguards will need to be calibrated to protect freedom of expression.

The objection of course is that this should not be the function of the Court but rather the Member States. Indeed, the same argument could be made concerning the Court’s filling of the policy gap by reading upload filters into Article 17 generally. In short, we are where we are because of the Court’s ruling – and we now face a dilemma. Article 17 envisages that legitimate content should not be blocked. The Court reminds us that a system that cannot sufficiently distinguish legal from illegal content will not comply with Article 11 of the Charter. Yet, the remaining safeguards within the provision deal with this precise scenario, giving rise to questions that put into sharp focus the lack of guidance from the Court as to what the substantive essence of freedom of expression is, in this context. To this end, even with the Court attempting to limit the overall scope of such filtering, mistakes will happen. What then of redress?

What if Time is of the ‘Essence’ of Freedom of Expression under the CDSM?

Let us assume the content was legal but erroneously blocked – how long is too long for the block to be disproportionate in a scenario where there are apparent safeguards so that this should not have occurred at all? Does it matter? After all, a safeguard is not a guarantee per se and in this context if it were, imbalance would surely occur in suggesting blocks could never occur. But this is precisely the point – if under classic proportionality balancing a block can occur, the redress mechanism must still be effective and, as Article 17(9) provides, without undue delay and subject to human review. The irony is that the latter, unable to achieve the former on sufficient grounds under Article 14 of the eCommerce Directive, is one of the main arguments in favour of upload filters. What then is a proportionate delay where the most appropriate redress is removal of the block? The answer may well prove elusive with the task being to apply objective standards to expressions that are by their nature inherently subjective and relative to the context in which they are made.

Indeed, coming within an exception or limitation only tells us the category of the expression – it says nothing as to the value of that expression to the user relative to any myriad of factors, from the imprint of user personality to the temporal resonance with society and culture at the point of upload. When it comes to freedom of expression in this context, time and not the existence of safeguards may well be the starting point for discerning the ‘essence’ of the right. Indeed, as we move towards greater filtered futures, we need a surer understanding of the ‘essence’ of freedom of expression in this context beyond the existence of safeguards. For the implementation of Article 17 in light of the Court’s ruling, ex ante mechanisms that allow prior user input to challenge any potential block appear a necessary minimal mitigation measure to protect internet user rights. But even this delays the inevitable question where a block is imposed on what turns out to be legal content – to what degree is time of the essence?

If the normative position following the Court’s ruling is that Article 17 is to be implemented, the assumption is that some delay is inevitable in resolving an end user’s complaint. If we accept the essence of freedom of expression is contingent on the existence of safeguards that include redress then this is unproblematic. However, in balancing copyright and freedom of expression in this context, the most proportionate (and appropriate) redress is the expeditious removal of the block, relative to the factors described above. If an objective standard remains elusive in this context precisely because of those factors, can we find balance? The perspective of classic fundamental rights balancing would still suggest an answer in the affirmative; after all, an end-user can still avail of out of court settlement procedures and the courts. However, if it is accepted that in this context the ‘essence’ of the right comes down to individualistic, nuanced and relative factors, such avenues appear wholly inadequate where time is of the essence. Following the Court’s ruling, the question of time may be derided as an abstract consideration that sits in contrast to the normative position of implementation. However, on the contrary, it is precisely because of the Court’s ruling that difficult questions remain surrounding the balance between freedom of expression and copyright in this context. In protecting end-users, a useful starting point would be to identify what the real ‘essence’ of freedom of expression is, in this context. If, as the foregoing suggests, time is of the essence, this invites the question – is balance in practice, on these terms, possible under Article 17?

The post Poland, the CDSM and the Court of Justice: Still Searching for the ‘Essence’ of the Fundamental Right to Freedom of Expression appeared first on COMMUNIA Association.

Article 17 of Directive (EU) 2019/790 on copyright in the Digital Single Market (CDSMD) has been subject to much debate even before its enactment. The latest twist is the CJEU’s ruling in the Polish action for annulment of Article 17 CDSMD. Uncertainties about the precise and correct practical application of Article 17 CDSMD remain. The judgment, however, provides some clarity on how this norm must be transposed into national law to ensure compliance with fundamental rights, particularly with freedom of expression and information as enshrined in Article 17(2) of the Charter of Fundamental Rights of the European Union.

Compatibility of verbatim transpositions and why the German approach is ahead of the pack

Article 17 CDSMD is open to various interpretations – as has become clear during the hearings before the CJEU. While Spain and France argued that an implementation of ex post safeguards is sufficient to protect user rights, the CJEU later confirmed the position taken by the Advocate General and Member States like Germany that Article 17 CDSMD’s ex post safeguards must necessarily be supplemented by ex ante safeguards. These should address the danger of overblocking, that is the undue blocking of lawful content by OCSSPs before its dissemination in order to comply with the obligations from Article 17(4) CDSMD.

The judgment emphasises the need for ex ante safeguards against rampant blocking under Article 17(4) CDSMD. This fact, together with the obligation for Member States, when transposing Article 17 CDSMD, to strike a fair balance between the various fundamental rights, have raised doubts about the compatibility of verbatim transpositions. Other commentators have rejected these, arguing that minimal verbatim transposition is necessary to avoid impairing the harmonisation effect of the Directive.

The CJEU did not concern itself with national transpositions, but rather solely Article 17 CDSMD in its original version, and found an interpretation in compatibility with the freedom of expression. The CJEU held that Article 17(4) CDSMD is accompanied by appropriate safeguards. The judgment requires Member States to ensure an interpretation of national provisions that contains these safeguards.

As the judgment itself already identifies an interpretation of the Article in line with fundamental rights, the same must surely also apply to identical wording (i.e., copy and paste transpositions) in national law. This has to be the case as Member States are bound to interpret their laws in line with the CJEU’s interpretation. National courts, when interpreting national law, must have regard to the case law of the CJEU. Therefore, a conforming interpretation of verbatim transpositions should be ensured. As a consequence, copy and paste transpositions must be considered compatible with the judgment and the fundamental right to freedom of expression.

This does not mean, however, that this kind of implementation is the best in the face of freedom of expression and information. Instead, a more elaborative implementation, which provides more details on the delineation of permitted and prohibited ex ante blocking, should be the preferable way forward.

A conceivably elaborative implementation is the German version of Article 17 CDSMD, as it was transposed in the act on the copyright liability of OCSSPs (UrhDaG). One aspect of particular interest is the concept of a “presumed legal use”. In summary, Germany established a national additional ex ante safeguard for content which either qualifies as minor usage or is marked by the user as legally permitted. Under certain requirements, this content is presumed to be lawful and therefore cannot be blocked by automated means implemented by the OCSSPs. If rightholders contest this content, they have to initiate the complaint procedure, which may result in the content being taken down.

While there is an ongoing discussion about the compatibility of the German mechanism with the EU template, it is true that it dares to do something that had been missing from the EU Directive: it defines circumstances under which ex ante blocking is not possible.

The need for a definition of “manifestly infringing”

It has to be said that while this constitutes a step in the right direction, the current German provisions may not be the ultimate solution. Rightholders argue that even the unjustified usage of a film sequence as short as 15 seconds can significantly harm their economic interests, when only blocked after an ex post intervention. Nevertheless, the German transposition puts requirements in black and white for the design and use of automated content recognition (ACR) technology and automated blocking based on it.

In order to protect freedom of expression, it is important to be more specific about the requirements and circumstances under which automated ex ante blocking of content is permissible. One of the key points from the judgment in the Polish case is that for content to be blocked ex ante without freedom of expression being unjustifiably harmed, no independent assessment of its unlawfulness must be necessary. In other words, content needs to be “manifestly infringing”, which makes this term the central standard for determining whether the prevention of an upload was lawful or not.

Therefore, it should not be left to OCSSPs to determine when content is infringing enough to be regarded as manifestly infringing and can thus lawfully be blocked automatically. Rather, regulators should find ways to define requirements. This would not only provide clarity to users, rightholders and platforms, but deployed at the EU level it would also contribute to the harmonisation objective.

The implementation of the German legislator may serve as a starting point. However, it only defines circumstances under which automated blocking is not permissible, i.e., when manifestly infringing content is not present. Therefore, the law only gives a hint of a negative definition. A positive definition, which indicates when content can be blocked automatically, has yet to be found.

Implications of the judgment for the design of the complaint mechanism

In the context of national transpositions of Article 17 CDSMD, two remarks regarding current questions of implementation should be made.

The first concerns national provisions in respect of the complaint mechanism as set out in Article 17(9) CDSMD. From the judgment in the Polish case, we know that the complaint mechanism is considered as an additional (ex post) safeguard, which applies in “cases where, notwithstanding the [ex ante] safeguards […], the providers of those services nonetheless erroneously or unjustifiably block lawful content” (para 93).

The complaint mechanism is therefore intended to deal with cases where there is a dispute as to whether the content is manifestly infringing. In those cases, however, it is in the nature of things that the content in question stays offline for the duration of the complaint mechanism. This presupposes that the basic requirements for ex ante safeguards have been implemented and that the ex post complaint mechanism only applies in exceptional cases. It is only under these conditions that provisions like the Italian one under which all contested content shall remain disabled for the duration of the complaint procedure can be considered compatible with the judgment.

The Commission’s category of “earmarked content” needs revision

The second aspect relates to earmarked content as mentioned in the Commission’s Guidance on Article 17. The Guidance defines earmarked content as content flagged by rightholders that is particularly valuable and could cause them significant harm if it remains available without authorization (examples include pre-released music or films). According to the Guidance, earmarked content should be specifically taken into account when assessing whether OCSSPs have made their best efforts to ensure the unavailability of content.

What is highly problematic about this provision, however, is that OCSSPs would be forced to exercise particular care and diligence in this case, which would ultimately result in a higher blocking rate and ignore the requirements of the judgment in the Polish annulment action. As a solution, the Commission presents rapid ex ante human review in the Guidance, which takes place for such earmarked content before the content gets online, when detected by the filters.

This, however, does not comply with Article 17(8) CDSMD and what follows from the Glawischnig-Piesczek and recent Poland cases. According to these cases, a provider can only be required to remove content where a detailed legal examination is not necessary. And, although framed as “rapid ex ante review”, this is nothing other than a detailed legal examination.

Therefore, the Commission needs to revise its Guidance on this point and Member States should choose an implementation of earmarked content which respects the case law. A possible solution could be to use an earmark mechanism not ex ante but ex post. Content which is marked by rightholders as of significant economic value and matches content uploaded by users could be processed through an accelerated complaint procedure. This would be similar to what Article 19 of the Digital Services Act (DSA) establishes.

The DSA needs to fix Article 17 CDSMD

The DSA raises hopes for more harmonisation of details related to the interpretation of Article 17 CDSMD. Due to a largely overlapping scope of application for OCSSPs in the area of copyright, it can be assumed that the DSA provisions apply on the basis of a lex generalis relationship to Article 17 CDSMD. Provisions such as Article 17 DSA, that sets out detailed rules for a complaint mechanism, or Article 19 DSA, with its trusted flagger regime, could influence the way Article 17 CDSMD works in practice.

Due to its nature as a regulation, the DSA should lead to greater harmonisation. In order to achieve this, it would in addition be necessary to use the aforementioned revision of the Guidance to develop a positive definition of manifestly infringing content which can be used as a basis for designing the algorithms of OCSSPs.

The post Implications from C-401/19 for National Transpositions in the Light of Freedom of Expression appeared first on COMMUNIA Association.

The adoption of the CDSM Directive marked several turning points in EU copyright law. Chief amongst them is the departure from the established liability exemption regime for online content-sharing service providers (OCSSPs), a type of platform that (at the time) was singled out from the broader category of information society service providers regulated for the last 20 years by the E-Commerce Directive. To address a very particular problem – the value gap – Article 17 of the CDSM Directive changed (arguably, after the ruling in (YouTube/Cyando, C-682/18) the scope of existing exclusive rights, introduced new obligations for OCSSPs, and provided a suite of safeguards to ensure that the (fundamental) rights of users would be respected. The attempt to square this triangle resulted in a monstrosity of a provision. In wise anticipation of the difficulties Member States would face in implementing, and OCSSPs in operationalizing Article 17, the provision itself foresees a series of stakeholder dialogues, which were held in 2019 and 2020. Simultaneously, the drama was building up with a challenge launched by the Republic of Poland against important parts of Article 17. Following the conclusion of all these processes, and while (some) Member States are considering reasonable ways to transpose Article 17 into their national laws, it is time to take stock and to look ahead. The ruling in Poland v Parliament and Council (C-401/19) is a good starting point for such an exercise.

Shortly after the CDSM Directive was adopted, the Republic of Poland sought to annul those parts of Article 17 which it argued, and the Court later confirmed, effectively require OCSSPs to prevent (i.e. filter and block) user uploads. Preventing control of user uploads constitutes, the Court confirmed, a limitation of the right to freedom of expression as protected under Article 11 of the EU Charter of Fundamental Rights. In its argumentation Poland suggested that it might not be possible to cut up Article 17, a provision of ten lengthy paragraphs. Indeed, the intricate relations between specific obligations, new liability structures and substantive and procedural user safeguards cannot be seen in isolation, and therefore require a global assessment. And this is what the Court embarked upon. Whilst assessing the constitutionality of Article 17, the Luxembourg judges, in passing, provided some valuable insights into how a fundamental rights compliant transposition might look, but also left crucial questions unanswered.

The Court was very clear that any implementation of Article 17 – as a whole – must respect the various fundamental rights that are affected. However, the judges in Luxembourg did not go into detail on how Member States should transpose the provision. Of course, it is part of the nature of directives that Member States have a certain margin of discretion how the objectives pursued by a directive will be achieved. To complicate matters, the Court stated expressly that the limitation on the exercise of the right to freedom of expression contained in Article 17 was formulated in a ‘sufficiently open’ way so as to keep pace with technological developments. Together with the intricate structure of Article 17 itself, this openness tasks Member States with a difficult exercise: to arrive at a transposition of Article 17 (and of course also other provisions of the CDSM Directive) that achieves the objective pursued while respecting the various fundamental rights affected (see Geiger/Jütte). The various national implementation processes have already demonstrated that opinions differ as to what constitutes a fundamental rights compliant implementation, and proposals have been made at different points of the spectrum between rightsholder-friendly and user-friendly transpositions.

The Court’s ruling makes a few, very important statements in this respect and thereby sets the guardrails beyond which national transpositions should not venture. First and foremost, control of user uploads must be strictly targeted at unlawful uses without affecting lawful uses. Recognizing that the employment of online filters is necessary to ensure the effective protection of intellectual property rights, the ruling highlights the various safeguards that ensure that the limitation of the right to freedom of expression is a proportionate one. Implicit here is that the targeted filtering of user uploads is a limitation of this fundamental right that necessitates six distinct measures that must be put in place, and which, in concert, ensure respect for the rights of users on OCSSP platforms.

Targeted Filtering

To limit the negative effects of overblocking and overzealous enforcement, any intervention by OCSSPs must be aimed at bringing the infringement to an end and should not interfere with the rights of other users to access information on such services. This suggests that it must be clear that only unlawful, or infringing content is targeted. The Court elaborates that within the context of Article 17 only filters which can adequately distinguish between lawful and unlawful content, without requiring OCSSPs to make a separate legal assessment, are appropriate. This is problematic, since copyright infringements are context sensitive, in particular in relation to potentially applicable exceptions and limitations. Requiring rightholders to obtain court-ordered injunctions before content can be subject to preventive filtering and blocking seems unreasonable, considering the amount of information uploaded onto online platforms. On the one side, copyright infringements are different from instances of defamation or other offensive speech, that was the subject of the preliminary reference in Glawischnig-Piesczek (C-18/18). On the other, the requirement of targeted filtering seems to eliminate the proposal made by the Commission in its Guidance (see Reda/Keller) to allow rightholders to earmark commercially sensitive content, which could be subject to preventive filtering for a certain limited amount of time. Unfortunately, the Court does not elaborate further how OCSSPs can target their interventions. It merely states that rightholders must provide OCSSPs with the relevant and necessary information on unlawful content, which failure to remove would trigger liability under Article 17(4) (b) and (c). What this information must contain remains unclear. Arguably, rightholders must make it very clear that certain uploads are indeed infringing.

User Safeguards

In terms of substantive safeguards, Article 17 takes a frugal approach. To avoid speech being unduly limited, it requires Member States to ensure that certain copyright exceptions must be available to users of OCSSPs. These exceptions are already contained in Article 5 of the Information Society Directive as optional measures, but Article 17(7) makes them mandatory (see Jütte/Priora). Arguably not much changes with the introduction of existing exceptions (now in mandatory form). Moreover, the danger of context-insensitive automated filtering still persists, even though users enjoy these substantive rights.

Therefore, Article 17 foresees procedural safeguards, which is where the balance in Article 17 is struck. With the Court having confirmed that preventive filtering is an extreme limitation of freedom of expression, the importance of the procedural safeguards cannot be overstated (see Geiger/Jütte). The ruling in its relevant parts can be described as anticlimactic. Instead of describing how effective user safeguards must be designed, the judgment merely underlines that these safeguards must be implemented in a way that ensures a fair balance between fundamental rights. How such balance can be achieved was demonstrated in summer 2022, when the Digital Services Act (which is still to be formally adopted) took shape. A horizontally applicable regulation that amends the E-Commerce Directive, the DSA puts flesh to the bones that the CDSM Directive so clumsily constructed into a normative skeleton.

The DSA provides far more detailed procedural safeguards compared to the CDSM Directive. It is also lex posterior to the latter, which in itself is, however, lex generalis to the former. Their relation, but arguably also their genesis, holds the key to outlining not necessarily the solution to the CDSM conundrum, but to the questions that national legislators must ask.

The DSA sets out how hosting services and online platforms must react to notifications of unlawful content and how they must handle complaints internally; the DSA also outlines a system for out-of-court dispute settlement, which requires certification by an external institution. Some of these elements are, in embryonic form as mere abstract obligations, already contained in the CDSM Directive’s Article 17. And by definition, OCSSPs are hosting providers and online platforms in the parlance of the DSA, which is why these rules should also apply to them. OCSSPs, however, incur special obligations and are exempted from the general liability of the E-Commerce Directive and (soon) the DSA, because they are more disruptive – of the use of works and other subject matter protected by copyright and of the rights of users, the latter as a result of obligations incurred because of the former.

That some of the rules introduced by the DSA must also apply to OCSSPs has been argued elsewhere (see Quintais/Schwemer). It has been suggested that the rules of the DSA should apply in areas in which the DSA leaves Member States a margin of discretion or where the CDSM Directive is silent. CDSM rules that derogate from those of the DSA, specifically the absence of a liability exemption for user uploaded content will certainly remain unaffected. But there are good arguments to be made why in areas of overlapping scope, the DSA should prevail, or systematically supplement the CDSM Directive. Instead, the DSA rules must form the floor of safeguards that Member States have to provide, and which should be elevated in relation to the activities of OCSSPs. The reason is a shifting of balance between the fundamental rights concerns, which relates to the last paragraph of the CJEU’s ruling in Poland v Parliament and Council. The obligation to proactively participate in the enforcement of copyright intensifies the intervention of OCSSPs, as opposed to the merely reactive intervention required under the rules of the DSA. The effects for rightholders are beneficial (although the rationale for Article 17 suggests that it addresses a technological injustice) in the sense that OCSSPs must intervene in a higher volume of cases; the negative effects are borne by users of platforms, their rights are limited as a result. Arguably, this must be balanced by a higher level of protection of users, in this case in the form of stronger and more robust procedural safeguards. A further elevation of substantive safeguards would in itself not be helpful, since their enjoyment relies effectively on procedural support.

As a result, Member States should, or even must, consider that the elaboration of user safeguards in the form of internal complaints mechanisms and out-of-court dispute settlement mechanisms must find concrete expression in their national transpositions. They should be more robust than those provided by the DSA. Ideally, this robustness will be written into national laws and not left to be defined by OCSSPs as executors of the indecisiveness of legislators. The difficulty lies, of course, in the uncertainty of technological process, the development of user behaviour and the rise and fall of platforms and their business models. Admittedly, some sort of flexibility is necessary, the Court has recognized this explicitly. But if the guardrails that guarantee compliance with fundamental rights are not, and possibly cannot be written into the law, they must be determined by another independent institution. One institution, understood more broadly, could be the stakeholder dialogues required pursuant to Article 17(10) CDSM Directive. The Court listed them as one of the applicable safeguards, and a continuous dialogue between OCSSPs, rightholders and users could serve to define these guardrails more concretely. Such a trilogue, however, must take as its point of departure the ruling in C-401/19 and learn from the misguided first round of stakeholder dialogues. Another institution that could progressively develop standards for user safeguards are formal independent institutions, such as the Digital Service Coordinators (DSCs) required under the DSA. There, they have the task, amongst others, of certifying out-of-court dispute settlement institutions and awarding the status of trusted flaggers. Their tasks could also include general supervision and auditing of OCSSPs with regard to their obligations arising not only under the DSA, but also under the CDSM Directive. In the context of the Directive, DSCs could also be tasked with developing and supervising a framework for rightholders as ‘trusted flaggers’ on online content-sharing platforms. The role of DSCs under the DSA and their potential role in relation to OCSSPs is still unclear. But the reluctance of the legislature to give substance to safeguards and to clarify the relation between the DSA and the CDSM Directive mandates that the delicate task of reconciling the reasonable interests of rightholders and the equally important and vulnerable interest of users be managed by independent arbiters. Leaving this mitigation to platform-based complaint mechanism or independent dispute settlement institutions is an easy solution. A constitutionally sound approach would try to solve these fundamental conflicts at an earlier stage instead of making private operators the guardians of freedom of expression and other fundamental rights. While the Court of Justice of the European Union has not stated this expressly, its final reference to the importance of implementing and applying Article 17 in light of, and with respect to fundamental rights can be understood as a warning not to take the delegation of sovereign tasks too lightly.

The post Implementation Imperatives for Article 17 CDSM Directive appeared first on COMMUNIA Association.

On the other hand, we have the German approach that proposes to implement Article 17 via a separate “copyright-service-provider law” (“Urgeberrechts Diensteanbieter Gesetz”) that substantially departs from the language in an attempt to capture the structure and effet utile of the directive.

The German implementation proposal focuses on using the room for legislative discretion left by the directive to give practical meaning to the abstract requirements to protect user rights contained in the directive. It also adds measures aimed at ensuring that individual creators directly benefit from the new rules. As a result, the German implementation proposal is much closer to the legislative compromise struck by Article 17 than any of the more literal implementation proposals that have emerged so far.

Over the past few months, the German implementation proposal has come under intense pressure from exclusive rightsholders and some platforms who argue that the proposed approach does not adequately reflect the provisions of Article 17. Besides, rightsholders have also claimed that it violates national and international copyright law in multiple ways. A central argument of the opponents of the German implementation proposal is the claim that it strays too far from the text of the directive.

Given this background, it is interesting to see the first Austrian implementation proposal (that was circulated to stakeholders for feedback earlier this week) take a middle road between the two existing approaches. The Austrian implementation proposal does integrate the provisions from Article 17 directly into the text of the existing Austrian copyright act, thus deviating from the structure of Article 17, but mostly stays very close to the text of the directive. At the same time, it takes up key elements first introduced in the German approach: The non-waivable direct remuneration right for authors and performers, the protection of minor uses from automated filters, the ability for users to flag uploads as legitimate, and the ability for users’ organizations to act against platforms that engage in structural over-blocking. The result is a proposal that (similar to the German one) focuses on strengthening the position of creators and users, instead of leaving it up to platforms and large corporate rightsholders to set their own rules.

The Austrian proposal in more detail.

So let’s look at the Austrian proposal in more detail: Similar to the German proposal it introduces a direct remuneration right for authors and performers that will ensure that independent of existing contractual arrangements with publishers and other intermediaries, creators will be remunerated for the use of their works on platforms. As in the German proposal, this direct remuneration right can only be exercised via collective management organizations, which means that it will primarily benefit creators in sectors with existing collective management structures. In the German discussion this direct remuneration right has been strongly criticized by both platforms, who would prefer not to pay for obvious reasons) and by intermediary rightsholders, who prefer to control how much (or rather little) of their licensing revenue should go to the actual creators. Given that the need to make sure that creators benefit from the use of their works on platforms was the main argument for getting Article 17 in the first place, the fact that rightsholders are now trying to undermine the proposed direct remuneration right is more than a little bit hypocritical. 

On the other side of the balance, there are three main interventions that strengthen user rights. Where the German Ministry of Justice proposed a new “minor uses” exception that would ensure that minor uses of existing works cannot be automatically be blocked, the Austrian proposal takes a slightly different approach. It borrows the definition of “minor uses” from the German proposal (less than 20 seconds of audio or video, less than 1000 characters of text, images smaller than 250KB), but instead of making them subject to an exception, the Austrian proposal specifies in § 89a (1) that automatically blocking these “minor uses” without human review is not proportionate (and therefore platforms are not allowed to do this). This is an interesting use of the proportionality requirement from Article 17(5) of the directive that is based on the correct insight that measures (such as automated blocking) employed by platforms must be proportionate concerning all the objectives contained in Article 17 including the requirement not to block legitimate uses in 17(7) and 17(9). In the explanatory memorandum, the Ministry notes that “It is precisely the use of such small excerpts that will often be covered by an exception or limitation and thus be permitted”. While this may be a less controversial approach, authors and performers would be better off in the German proposal: By legalizing minor uses under an exception, the German proposal would create the possibility that they receive compensation for those uses.

Excluding minor use from automated filtering is one of two measures to ensure that the Austrian proposal complies with Article 17(7). In addition to this, the proposal also requires platforms to provide users with the ability to flag uploads as covered by an exception. § 89b (7) specifies that platforms must display uploads flagged by their uses as long as the legal nature of the use is obvious to the platform. While welcome in principle, this flagging mechanism is more limited when compared to the mechanism in the German proposal: It only applies to uses under the exceptions noted in Article 17(7) of the directive (with the notable omission of use for the purpose of citation) while it should apply to uses under all relevant exceptions. Besides, the flagging mechanism can’t be invoked if a work is used under open licenses or is not protected by copyright at all.

The third measure to protect user rights is the right for users’ organizations to file cases against platform operators that systematically fail to meet their obligations to protect legitimate uses. § 89b (7) allows user organizations to bring cases against platforms that systematically over-block user uploads in order to stop them from doing so. As noted in the explanatory memorandum the ministry sees this as an important additional safeguard to protect users against violations of their fundamental right to freedom of expression. The ability to bring collective action against platforms is further strengthened by an explicit requirement in § 89b (7) for platforms to provide users and user organizations with information on the functioning of any measures they employ to comply with Article 17. Taken together these two measures constitute an important check against over-blocking that is missing from all other implementation proposals. While the German proposal does contain the ability for users’ organizations to take platforms to court over overblocking, it lacks the necessary information rights over the functioning of blocking measures and the remedies to bring about systematic changes in the way that platforms police copyright-protected content.

A template for implementation?

In general, the Austrian example appears to fuse together some of the explicit attempts of the German Ministry of Justice to implement the directive in a way that minimizes the over-blocking of legal content, provides strong safeguards for user rights and ensures that authors and creators are directly remunerated, on the one hand, with the desire to stick as closely as possible to the text of the directive, on the other hand. The result may be the most realistic implementation proposal yet.

While the elements that we have outlined above will almost certainly come under attack from large corporate rightsholders (and possibly from the platforms as well), it is important to recall that, according to the Commission’s guidance proposal, it is simply not enough to do a literal implementation of the directive that does not specify concrete safeguards against automated blocking of legitimate content. Having closely watched the discussion on the European level and in Germany, the Austrian Ministry of Justice may just have shown us (and more importantly the other Member States still waiting to propose their implementations) a credible way for reconciling the conflicting requirements of Article 17 of the DSM directive.

The post Austrian Article 17 proposal: The high road towards implementation? appeared first on COMMUNIA Association.

In order to achieve this objective, the German discussion draft published in June relies on the idea of “pre-flagging”: users would be allowed to flag uploads containing third party works as legitimate. Platforms would then be prevented from automatically blocking pre-flagged uploads unless they determine that the flag is incorrect because the upload is “obviously infringing”. 

By contrast, the Commission’s implementation guidance consultation proposes a “match-and-flag” mechanism: if upload filters detect the presence of a third party work in an upload and the use is not deemed to be “likely infringing”, then the uploader is notified and given the ability to state that the use is legitimate. If the user flags the upload as legitimate, the platform will have to initiate a human review of the upload, which remains available from the moment of upload until the review has been concluded. This type of mechanism was first suggested by a group of copyright academics in October of last year. It is also at the core of the proposal that we had presented during the last meeting of the stakeholder dialogue.

Both approaches provide a mechanism that limits the application of fully automated upload filters (while implicitly acknowledging the fact that many platforms will deploy upload filters). In the Commission’s proposal, filters are limited to making a pre-selection (“is the upload likely infringing?”); in the German proposal, they can only operate on unflagged content and to filter out “obviously incorrect” pre-flags.

Convergence on “match-and-flag”?

Both approaches have been criticised by rightholders, who claim that they undermine the “original objective of the directive” without providing alternative proposals on how automated filtering can be reconciled with the requirement not to block legitimate uploads. In addition, the German discussion proposal has also been criticised by platforms such as Google and Facebook. The platforms are arguing that giving users the ability to pre-flag every single upload would be impractical and would likely lead to substantial numbers of unnecessary (where the content in question is already licensed) or unjustified (users making excessive use of the pre-flagging tool) pre-flags, which would make such a system impractical to operate at scale. 

Netzpolitik.org has now published a leak of a new version (“Referentenentwurf”) of the German implementation law proposal. This version abandons the pre-flagging mechanism and replaces it with a “match-and-flag” approach similar to what the Commission has proposed (it also closely resembles a suggestion made by Google in its response to the German Consultation). However, there are also important differences between the two proposals, and based on a closer analysis it is clear that the new German proposal offers considerably less protection against unjustified blocking or removal of uploads than either the initial pre-flagging approach or the approach proposed by the Commission. To understand why we need to look at the details of the proposed mechanisms.

Both approaches clearly assume that platforms are able to identify matches between uploads and works that rightholder have requested to be blocked in (near) real time. Both the Commission’s proposal and Article § 8 of the German Referentenentwurf assume that users can be notified of a match during the upload process and thus can prevent legitimate uploads from being blocked at upload. While some technology vendors claim to have the ability to reliably match content during the upload, it is currently unclear if the ability to match in (near) real time is widely available to all platforms.

Given the uncertainty about the availability of real-time matching solutions for all types and sizes of platforms, it must be ensured that the use of automated filters is not imposed de facto by national legislators if this could be disproportionate for smaller platforms. The New German proposal does seem to require the use of real-time filters which would make it incompatible with the proportionality requirements in Art 17(5). 

The limits of “match-and-flag”

But even if we assume that platforms have the ability to match in real time during the upload, the approach still has limitations. The requirement to make best efforts to prevent the availability of works in Article 17(4)b does not apply only to new uploads: it also applies to uploads that are already on a platform. In situations where rightholders provide platforms with new blocking requests, the platforms will need to make best efforts to identify and remove them as well (this problem will be especially acute at the moment when the directive comes into force). Notifying the uploader of a match and giving her the possibility to flag the upload as legitimate does not offer the same protection here, because it cannot be assumed that the user has the ability to react immediately. This would mean that the upload in question would become unavailable until the uploader has had a chance to object. 

This problem is much more pronounced in the new German proposal. The Commission’s proposal makes it clear that platforms are only allowed to automatically remove uploads if a match is “likely infringing”. This means that already uploaded works that do not meet this requirement cannot be removed until either the user has had a chance to react to a notification or until the platform has concluded a human review of the upload in question. The German proposal does not contain such a safeguard, as it requires the automated removal of uploads unless these have been flagged as legitimate during the upload. 

This is regardless of whether the match is likely to be infringing or not. In situations where users cannot react to notifications right away, this will result in the removal of substantial amounts of legitimate uploads. Under the previous German pre-flagging mechanism this would not be an issue (with the exception of uploads already on the platform when the German implementation enters into force), because users would have had the ability to flag any legitimate upload as legitimate. The new German proposal only gives them the possibility to flag works as legitimate that are already on a blocklist at the moment of upload. 

Towards a combined approach?

As long as this blindspot persists, the new German proposal does not adequately implement the requirement in Article 17(7) that the availability of uploads that do not infringe copyright must not be prevented by measures deployed to implement Article 17(4)b. To fix this, the German legislator should add, to the mechanism provided in §8 of the new proposal, the ability to flag any upload as legitimate after it has been uploaded and that flagged upload cannot be automatically blocked. 

This combined approach would provide even stronger safeguards than the Commission’s proposal, which hinges on the idea that it is possible to automatically differentiate between likely infringing and likely legitimate content based on technical parameters.

As we have pointed out in our response to the Commission’s consultation, this approach, while viable in principle, is flawed as long as defining those technical parameters is left to platforms and rightholders without any involvement from users’ organisations. In addition, the proposed “likely infringing” standard does not set a high enough bar for preventing automated removal of potentially legitimate content. Instead, the “identical or equivalent” standard proposed in the academic statement that introduced the idea of “match-and-flag” should be a point of departure. In the case of time based media, this could be operationalised as matches that are at least 20 seconds long and where the match consists of at least 90% of the original work and at least 90% of the upload in question. In addition, matches of indivisible works (such as pictures) and short works (such as short poems) should never be assumed to be infringing, even when they correspond to 100% of an upload.

Meaningful protection for Public Domain and openly licensed works

A final advantage of such a combined approach is that it would also offer real protection from automated blocking for works that are in the public domain or available under open licenses. While such works are free to use for anybody, they are frequently blocked or removed as the result of wrongful ownership claims. In this situation it must be possible for anyone at any time to flag such works as being in the public domain or openly licensed. Given that this status will be the same across all (types of) platforms, such flags should not be recorded by individual platforms but in a public database that must be consulted by any system as part of assessing the status of an upload. 

While it may make sense for platforms to use their own private databases when it comes to matching uploads to reference files of works to be blocked, the effective protection of public domain and openly licensed works requires a fully transparent public database that reflects their status as public goods. This public database must be consulted by any system as part of assessing the status of an upload and should be maintained by an independent trusted entity that also offers a conflict resolution mechanism for resolving conflicting claims. 

Summary

At this stage, there seems to be some level of convergence towards “match-and-flag” mechanisms as a practical approach to reconciling 17(4) and 17(7). While still exhibiting shortcomings, such an approach would reflect the internal balance of Article 17 that the EU legislator arrived at. In how far a “match-and-flag” mechanism will be able to put this balance into practice depends on its practical implementation. As we have outlined above this means that: 

• There must be high thresholds to presume infringement and consequently permit fully automated blocking of uploads.
• These thresholds should be based on fully transparent criteria, which users can challenge in court.
• All matched uploads that do not meet these thresholds must be protected from blocking and flagged uploads must not be removed while under review by the platform.
• In addition there must be the ability for anyone to pre-flag works that are in the Public Domain or available under an open license via a decentralised public database that must be consulted by any (automated) measures used to comply with Article 17(4).
• National implementation must contain safeguards that ensure that already existing uploads cannot be blocked automatically.

Finally, it must also be ensured that the use of automated filters is not imposed by national legislators if this would be disproportionate for the platform in question.

The post Taming the upload filters: Pre-flagging vs. match and flag appeared first on COMMUNIA Association.