A step ahead toward filtering over large social platforms?

 

Schermata 2019-10-03 alle 11.17.49

Facebook may be ordered, de facto, to monitor the activity of its users in order to prevent them to repeat posting identical or equivalent unlawful content. This is what the European court ruled today in a judgment referred by an Austrian court (Judgment in Case C-18/18 Eva Glawischnig-Piesczek v Facebook Ireland Limited) and concerning the request of a politician to remove various  comments and allegations published by a Facebook user which were harmful to her reputation. Since such comments and allegations were identical or had an equivalent content, the question was whether Facebook, after removing a first time, should keep an eye on future similar illicit behaviors.

The European court ruled in the positive way. It could sounds like a common sense decision in the normal world, but in the case of a social network such an obligation would imply an automated system able to intercept such illicit content and evaluate whether they are identical and/or equivalent. Basically, an AI filter or something very sophisticated would be needed. This is not a little thing in the matter of social networks because a kind of censorship activity would be delegated to a machine.

Today’s European decision may create an important shift from consolidated interpretation of the liability regime of hosting providers under the current European framework, that is to say the European Electronic Commerce Directive (Directive 2000/31). The current system provides per the so-called “notice and take down” mechanism (article 14), whereby an host provider such as Facebook is not liable for stored information if it has no knowledge of its illegal nature or if it acts expeditiously to remove or to disable access to that information as soon as it becomes aware of it. With the news decision, the mechanism may turn into a kind of “notice and stay down”, whereby Facebook would be liable for illicit content not notified to him, provided that they are identical or equivalent to previous notified ones. Facebook should be de facto aware of such future posts, something which sounds a bit peculiar, unless you ask Facebook to constantly check what users are doing on the platform.

This is a huge development in terms of liability regime by hosting providers, because it may imply a de facto monitoring obligation, in contradiction with art. 15 of the same directive which instead affirms that no monitoring obligation may be imposed on hosting providers. Fact is the court reminds that “the directive prohibits any requirement for the host provider to monitor generally information which it stores or to seek actively facts or circumstances indicating illegal activity”. However, the current judgment seems to undermine in practice such an obligation, by stating that Facebook should:

– remove illicit content posted on its social network, the content of which is identical to the content of information which was previously declared to be unlawful, or to block access to that information, irrespective of who requested the storage of that information;

–  do the same when the new/future content is just “equivalent” to content previously declared unlawful.  The “equivalence” assessment could be done via automated search tools and technologies, since a human intervention does not seems feasible.

This is a paramount decision which may influence the activity of the incoming European Commission, which was already considering to revise the liability regime of online intermediary and platforms via the announced Digital Service Act. Others may even invoke this case in relation to the potential filters which could be imposed when transposing art. 17 of the recent Copyright Directive in the matter of video-sharing.

In fact, one would have expected more cautions by the European court in rendering a decision which may impact on the freedom of speech of Internet users, whose content could now be intercepted and blocked by machines. However, today’s decision seem to refer to a necessary previous assessment by a national court, which should lay down all the needed guarantee and limits to protect the freedom of citizens to share information and opinion on social network. Of course, a bit more guidance by the EU Court for the national judges would have been appreciated, and this lack may be a reason for a new preliminary rulings sooner or later.

This is why today’s decision should not be regarded as  leading case allowing filtering on platforms on generic way. The European court recognized that a fair balancing of interests is required, this is why the (filters) obligation can be imposed only by a judge which should consider, inter alia, the proportionality of the measure taking into account the offensive effect of the illicit content, the rights of users to express their opinion, the capability of the concerned platform to install a monitoring software ecc.

Interestingly, the European court does not preclude that an “a stay down” order could be applicable on worldwide basis, that is to say on all national versions of Facebook.

 

The new European Commission, a nightmare for dominant online platforms?

Schermata 2019-09-10 alle 12.40.16

The next European Commission may cause headache to dominant OTTs in Europe. In fact, the allocation of the portfolios announced today by the appointed President Von der Leyen reveals the intention to supervise and intervene in the competitiveness of digital sector even more effectively than in the past.

Firstly, Margaret Vestager, who in the last 5 years lead landmark cases against dominant online operators (Google, Apple, Amazon, Qualcomm ecc) got confirmed in the portfolio for competition. In addition, she was granted a Vice-president role for the entire digital sector, so as to be empowered and supervise all other commissioners operating in the digital area. This means that she will have even more effective powers in case she intends to take measures against dominant OTT. However, it should not be about higher fines (the current powers are already sufficient), while about the possibility to enact general rules targeting anticompetitive behaviors at origin, so that to make antitrust intervention superfluous. It could be about interoperability, transparency, conflict of interests ecc. .

Secondly, the digital agenda portfolio has been moved under the control of Sylvie Goulard,  who will also lead the Internal Market directorate. This means that a French representative will play a fundamental role in the likely revision of the e-commerce directive and in setting the future digital framework via the announced Digital Service Act. It is not a secret the France intends to find a way to address the predominance of US online operators in the European market. The appointment of a French personality for the Digital Agenda portfolio reminds the appointment of the German commissioner Oettinger in 2014 for the same position. That choice reflected the fear and the obsessions of the German government vis-à-vis Google. The most significant result was, at the end, the copyright directive, a very controversial legislation whose effects may however impact the Internet in general rather than dominant operators.  It will be interesting to see how the new digital commissioner will be able to distinguish the fight against dominant operators from the general rules of the Internet.

To sum up: dominant OTT will have to face a double threat: on one side VP Vestager will continue to exercise her ex post competition powers, on the other Commissioner Goulard may apply ex ante regulation. A scary scenario for US OTT, which may feel to be caught between two fires. 

One should consider whether this is a bad news just for dominant OTT, or for the Internet in general. Fact is, a French commissioner may be induced to propose restrictive rules for the internet, mirroring similar recent initiatives by the French government in the area of copyright, platform responsibility ecc . However, it is too early to foresee it, since commissioners are in principle independent and are normally expected to skip the desiderata of their government.

In addition  the supervising role of Margaret Vestager seems to be reassuring. The Danish politician has been as hard vis-à-vis dominant OTT as protective vis-à-vis small companies and individual rights. She will surely oppose regulations that, instead of targeting dominant operators, may affect the Internet in its whole. Her mission letter confirms that she will have to take care of  the digital eco-system in its entirety, in consideration also of SMEs.

 

The separation of telecom networks in Europe: from regulatory remedy to new business models for telecoms

structural-separation

The subject of how to curb the power of dominant companies in the communications sector has recently returned in vogue thanks to the intervention of US democrat Elizabeth Warren, who advocated a plan to downsize big online platforms (such as Google, Facebook, Amazon etc.) in order to contain their economic power. Should such a project were ever implemented, it would not be the first time in America, since a similar plan was pursued in the 80s with the dismantling of AT&T and the creation of the c.d. Baby Bells. At that time it was a gigantic operation, which brought down the value of At&T by 70% and completely reshaped the US telecommunications market (AT&T was not only the dominant telecom provider, but also a main manufacturer of equipments). While waiting to know how this debate will continue in the United States, let us examine how the issue of containing dominant operators in the communications sector has instead developed in the European perspective.

The European Union and the separation of telecom networks

In Europe this issue is essentially about telecoms (since big European OTT never existed so-far), and it is normally placed differently than in the United States: rather than downsizing economic dominance, the EU preferred to act against the vertical integration of telecom incumbents to reduce the effects of potential abuses and discrimination against competitors, namely new comers who were asking for access in order to compete in the retail market. Thus, the European telecom legislation of 2009 provided for a standard remedy on “functional” separation, aimed at separating wholesale functions and departments from the incumbent’s retail business, without however imposing a drastic structural separation of the network. Despite the high debate this remedy, inspired by the pre-existing model of separation of BT and OpenReach in the United Kingdom, never found significant application in the EU – with the exception of Italy, were incumbent TIM committed in 2008 a similar network organization. Fact is, the European Commission, represented at the time by the pugnacious Viviane Reding (the lady of both roaming and GDPR reforms) had to give up the main option, that is to say the real separation (structural) of the incumbent’s network via the creation of a distinct entity. This type of regulatory remedy found too much hostility by the larger European countries, particularly France and Germany. At the end, this type of regulated separation was included in the European frame work just as “voluntary”, that is to say as a unilateral act of the incumbent (normally in exchange for some deregulation of the sector). Even in latter case, however, there were no significant practical applications, apart from Sweden, where Telia separated the network in 2008 by creating the newco Skanova (a project which was however reconsidered a few years later). An operation of this kind is now underway in the United Kingdom, where OFCOM, thanks to its super-powers (far superior to those of other European regulators) has convinced BT to consent to the separation of Open Reach. The process is at an advanced stage but, remarkably, the separation only concerns the Open Reach staff and non-network assets, while the incumbent’s network remains the property of BT.

The new European Electronic Communications Code

The new European Code (adopted last December 2018 and which will be implemented by European member States by the end of 2020) did not provide for a regulatory remedy on structural separation of the dominant telecom networks, while the old rules on both functional and voluntary separation were confirmed. Compared to 2009, this time there was no battle either, since the subject of network separation by way of regulation had lost interest among European governments and regulators. The reason for this diminished interest lies in the fact that structural separation is a very difficult option to implement against hostile (and mostly listed) companies; it is in fact a question about separating business functions, rather then companies, therefore the boundaries between wholesale and retail are difficult to trace in the absence of friendly cooperation by the management.

A new model for the separation of telecom networks: the utility model

Meanwhile, the issue of the separation of telecom networks has is make its way in Europe in other forms. Having abandoned the idea of ​​a regulatory remedy, as in the case of structural separation, or negotiated deregulation, as in the case of voluntary separation, in some countries the separation of the network instead emerged as a business opportunity. In the Czech Republic and Denmark the local incumbents proceeded to spin off the network on the basis of financial considerations. By doing so, the separated telecom network is treated as a distinct asset within the entire telecom business of the incumbent, with the aim of making it more linear and interesting for investors: in fact a pure telecommunications network, without involvement with retail, appears similar to an utility business (such as energy, gas, transport, water) characterized by reasonable and certain returns of investments, defined in the long term. It is no coincidence that the two operators involved, respectively CETIN in the Czech Republic and TDC in Denmark, are both controlled by investment funds, that is, by investors who have a long experience and a favorable approach to the financing of utilities.

Separation of the telecom and wholesale-only network

This new course of the separation of the telecom networks would seem to find a backing in the new European Code of electronic communications, whose article 80 foresees a light regulatory regime for the “wholesale-only” operators, namely telcos exclusively installing and operating telecom networks for the benefit of other operators, and with no involvement in the retail business. The grant of such a light regulatory regime is based on the assumption that this type of operator is spontaneously encouraged to provide access services without discriminating or abusing against access seekers, which is in fact are seen just as clients and not as competitors (unlike the case of the vertical-integrated telecom network); furthermore, the Commission believes that this type of operator constitutes a natural vehicle for investment by infrastructure funds. Remarkably, art. 80 of the new European Code requires the separated network to be truly independent from the retail business, in other words the light regulatory regime can be invoked only when the incumbent has lost contro lover the network (which is not the case, yet, of CETIN and TDC).

The separation of telecom networks as the entry point of a new business model: retail services as a bundle of connectivity, services and content

Should the network separation model actually develop and expand (eventually with the wholesale-only variant) also thanks to growing interest by infrastructure investors, the telecom sector may begin to change dramatically. On one side, network operators may find convenient to consolidate massively, with the result of the return to the quasi-monopoly in some areas; on the other side, the retail market may become more dynamic, also with the entry of new players such as content producers or value-added service providers, which could result in offers of connectivity bundled with streaming or OTT services. Finally, telcos and content/service providers would face these mergers that have often been resolved in disasters in the past. If this scenario were actually realized, there would be plenty of new antitrust and regulatory implications, thus respective supervisory authorities should start even now to keep an eye on developments.

Network separation and 5G

The business of separating telecom networks could also be relevant in the 5G debate. Despite mobile operators being conservative in defending their vertically-integrated model, it is clear that the very high investment costs required for the 5G roll-out could make convenient the option of a single, not vertically integrated network. Such an idea was suggested some time ago with an internal memorandum of the Trump administration, which considered the single network option a possible remedy to stand up to the Chinese in the run for the 5G. The document was denied after protests by US operators, but nevertheless reflected real strategic discussions within the US administration. Of course, it is unlikely that a single and open 5G network could ever be imposed by way of legislation, but some government could still try to encourage it. Even if mobile operators tended to oppose it, they could never deny the evidence that the installation of a plurality of 5G networks is an unrealistic hypothesis. Remarkably, some of them are already thinking about network-sharing precisely to reduce costs (there are reports in Italy about TIM and Vodafone).

The European Copyright Directive/2: to filter or not to filter?

Schermata 2019-03-27 alle 12.09.53

[continuing from the previous post about the link-tax]

Art. 17 (former 13) of the Copyright Directive, concerning the liability of online content-sharing platform and the upload filters, is the most “systemic” part of the European copyright reform.

The primary liability of content-sharing platforms 

Article 17 creates a primary liability for online content-sharing service providers giving public access to copyright-protected works uploaded by their users. This is about Youtube or Youtube-like companies. Primary liability means that the platform is liable ipso facto when the illicit content is uploaded, because of the simple act by a third person (the platform user and uploader). It is a very burdensome liability operating ex ante, similar to the ones the civil law provides for very dangerous activities. To better explains this legislative innovation, please note that the current liability regime (in force since 2001 with Directive 2000/31) operates ex post, i.e. via the removal of the illicit content after the uploading. There is a huge difference between a) imposing a system preventing the upload of content ex ante, on one side; and b) providing an ex post removal regime, on the other; and legal implications are very different. If you catch this difference, you are looking at the core of the Copyright debate. All the rest is fresh water 🙂

As a consequence of the above “ipso facto” primary liability, content-sharing platforms must be authorized to have their content uploaded and made available on their website. If not, they are liable and should pay damages. But how can you be authorized even before users decide to upload a given video or music? You should foresee the future. This mechanism seems contrary to common sense and in fact it is.

How could content sharing platforms avoid primary liability?

Still, art. 17 provides that Youtube & CO. can avoid liability thanks to the so-called “mitigation measures”, consisting in demonstrating that they;

(a) made best efforts to obtain an authorisation, and

(b) made, in accordance with high industry standards of professional diligence, best efforts to ensure the unavailability of specific works and other subject matter for which the rightholders have provided the service providers with the relevant and necessary information; and in any event

(c) acted expeditiously, upon receiving a sufficiently substantiated notice from the rightholders, to disable access to, or to remove from, their websites the notified works or other subject matter, and made best efforts to prevent theirfuture uploads in accordance with point (b).

Filter or not filter?

Point b) is obscure. What does it mean “best efforts to ensure the unavailability of content on the platforms”?  Does this provision imply mandatory upload filters or not? This is a tricky. The new Copyright Directive never explicitly mentions filters and its supporters, both stakeholders and persons of the European institutions, have been vocal in declaring  that upload-filters are not foreseen by the provision. See for instance rapporteur Axel Voss:

Schermata 2019-03-27 alle 00.08.35 Sometimes

However, the sole available technological instrument that a platform can use to effectively avoid primary liability, that is to say to anticipate and prevent a user from uploading and making public illicit content, is a un automated upload-filter. Only filters can enable platforms to identify and block posting and -reposting of illicit content, before they become public.  No other technological instruments are available, since non-automated instruments, such as for instance a human check, would be unfeasible. Therefore, only two options are possible:

a) either we accept, even obtorto collo,  that only (automated) upload filters may preventively and effectively block illicit content, so that to give a sense, although controversial, to this Copyright Directive; or

b) we could be satisfied with any solutions other than automated upload filters. But this means that such solution could operate only ex post, after illicit content is uploaded and made public (similarly to the current liability regime). Tertium non datur.

Why this Directive pushes towards filtering, but with vague and imprecise terms?

Omitting the term “uploadfilter” in the Directive was a clear and intentional political decision, necessary  to facilitate the approval, this is why the supporters of the Directive have been denying that the same is mandating or implying upload filters. At the same time, the Directive itself requires platforms to make make “best efforts” to block illicit content, an objective which could be realistically achieved only by applying upload filters. In practice, the Directive aims at something specific which, however, cannot be said and written down. In the best scenario, it looks like the directive was written down by rabbits, not persons. In the worst scenario, it looks like an imbroglio.

National authorities to decide about filtering?

Since art. 17 is so vague, it will be a matter of national implementation and interpretation by national authorities. When looking at practical cases, national authorities will be able to take various directions but, at the end, only the same 2 options are available: either mandating upload filters, as the only effective instruments to prevent ex ante illicit content to be uploaded on platforms; or any other solutions which, however, will basically work on a ex-post control basis.

What about voluntarily filtering? This could be a possible solution but, as far as filters are not mandatory, they could be replaced by “best efforts” ex-post mechanisms. And, as we know, filters are a big investments that only big companies like BigG can afford. “Simple” filters could be found in the market but would be ineffective and overblocking, with the risk to kill the business.

What are national authorities expected to do in practice? If they look at the text of article 17, imposing an express upload-filters provision will not be obvious, also considering that the Directive contains another contradicting obligation (art. 17(8)): “The application of this Article shall not lead to any general monitoring obligation“.

Rights-holders will probably complain because they pretend unauthorised content to be effectively blocked ex ante, but then they also should explain why upload filters were not required before voting of the directive, while after the approval they are.

Anyway, there might be national governments which already have clear ideas. The French government, for example, just the day after the approval of the Directive officially stated that they are going to adopt upload filers because they are the only way to make art. 17 working:

Schermata 2019-03-28 alle 18.04.50

Will the Commission recommend filtering?

Unless you are French, it will be difficult for national authorities to take a straight decision on the topic. It is more likely that vagueness and contradictions of the Directive will be transposed into national law, without solving the problem which, therefore, will have to be dealt on the ground: at the very hand, a national judge will have to take the decision on the basis of vague provisions and therefore jurisprudence could strongly differ from country to country. A preliminary ruling to the European court of justice could also be possible.

In the meanwhile, the European Commission will be required to provide guidance on the matter after having heard the views of all concerned stakeholders (platforms, rights holders, users, ecc.) gathering together (art. 17(10)). It is unlikely that such stakeholders will magically find a an agreement on the filters subject, therefore at the very end the Commission will remain holding the bag and will have to decide what to do, what to recommend. Remarkably, the entire system seems designed to postpone, as longest as possible, any decision.

Worth-noting, next European Parliament and Commissioners may be a bit colder vis-à-vis copyright supporters. Therefore, will next Commission recommend national authorities to mandate upload filters or could, instead, recommend a more soft interpretation? This could probably become a question that the new European Parliament will ask to the next candidate commissioner for the Digital Agenda, in July 2019.

 

The European copyright directive /1: ancillary copyright, link-tax or what?

IMG_1715

The new Copyright Directive have been approved by the European Assembly with an important but not large majority: 348 in favor, 274 against, 36 abstention. Quite an important number of MEP avoided the voting, probably because they fear the revenge by citizens at next European elections in May. While the PPE (christian democrats) voted massively in favor of the Directive, other political parties , in particular S&D (socialists) and ALDE (liberals) were basically split.

Schermata 2019-03-26 alle 15.51.25

The completion of the legislative process

The path for the approval and effective application of the directive will still be long, although the concerned provision cannot change anymore. Council is expected to endorse very soon the same text approved by the Parliament (likely in April 2019) which then will be published in the Official Journal. At the point the legislative process will be duly completed. However, being a directive, the text still will need to be implemented into national law, and Member States will have 24 months time. Therefore, the concerned provisions are expected come into force in mid-2021.  

Implementation will not be an easy matter: especially with regard to the most controversial provisions, such as article 11 (now 15 with the new numbering) about the so-called linktax, and article 13 about liability of videosharing platforms and use of uploadfilters (now art. 17 with the new numbering) are written sometimes in a very vague way. The vagueness is not result of technical incapacity, rather it is the consequence of the difficulties in finding an acceptable compromise for everyone. Bismark’s saying whereby “Laws are like sausages. It’s better not to see them being made” is still valid more than ever.

As a final result, the new Copyright Directive risks to be implemented, interpreted and applied in many way from country to country. European Guidelines will probably facilitate more harmonization, however they will not be binding for national authorities.

The ancillary copyright 

Schermata 2019-03-26 alle 17.10.58

According to Art. 15 of the Directive (former art. 11) publishers  are granted a new ancillary copyright (pursuant to the Copyright Directive 2001/29/EC) for the online use of their press publications by “information society service providers”. The ancillary right will last two years after the press publication is published (art. 15(4)).

Remarkably, art. 15 will apply to any Internet service provider, with no exceptions about turnover, size and scope of the activity ecc. Being Google, a start-up, a charity or or an educational entity will be the same: there is no exemption (there might be an exemption just for persons, see below). This is the reason why Wikipedia has been so vocal against the reform. Someone believed that Wikipedia was exempted, but that exemption concerns just videosharing, not publishing rights (see art. 2(6)). Even a porte-parole of the European Parliament got confused on this matter:

Schermata 2019-03-26 alle 23.32.05

Surprise: the ancillary right is still a link-tax

According to art. 15(1)3, the ancillary right should nor apply to acts of hyperlinking, neither upon individual words or very short extracts of press publications (art. 15(1)4). What does this mean? The ambiguity of the provision is the reason why the name “link-tax” remains valid despite the attempts to remove it. The clarification that hyperlinks are not taxed per se is good but insufficient, since the social and economic value of hyperlinks derive from combination with words and images referring to the linked content. Should such words be more than one (at least 2) or form an extract which is not considered “short”, then the link-tax (or whatever is the name) will apply. In particular, it will not be possible to express with the hyperlink the title of the linked press article.

The only way to be safe would be to use hyperlinks combined with words disconnected or irrelevant from the content they refer to. However, this is not the way hyperlinks and the Internet work.

The way this provision will be applied from country to country may strongly vary. This the reason why Wikimedia Italia is so worried. In an exchange of tweets with a journalist of La Stampa, their worries are pretty clear:

Schermata 2019-03-26 alle 19.57.47

Will normal people be exempted? 

According to art. 15(1)2, the ancillary right should not apply to private and non-commercial uses by individual users. This is commendable, however the factual application may be very limited, since pure “Private or non commercial use” in the Internet is just theoretical. Internet services (blog, social platforms) normally require users to accept a set of contractual conditions which include eventual economic exploitation of content (mainly via revenue sharing of online advertising). Therefore, in order to make the exception valid, all Internet services should re-write their Terms & Conditions in order to clearly separate “private/non commercial” from “commercial” usage (even when commercial is just potential). This separation, however, may be burdensome and incompatible with current business models and, so far, no-one made an assessment to verify whether it can be made without affecting normal usage.

In any case, Internet providers will not be obliged to create such separation amongst their subscribers, with the consequence that the latter will remain in a limbo: private or commercial use? Who knows.

Will Google finally pay?

A part from the above, the greatest paradox is that the real objective of this fight, namely Google, may easily avoid the payment. The level of the fee payable with link-tax can be negotiated or even waived, therefore Google, thanks to its economic strength in the online world, could easily get the better conditions, or even pay ZERO as it already happened in Germany. Smaller Internet provider will not have the same privilege, though. Of course, one could start an antitrust action against Google for abuse of dominant position (in Germany someone did it), however Google is not dominant in the news aggregation market, therefore the action will likely fail.

At least, journalists will be paid more?

According to art. 15(5), journalists should receive an “appropriate” share of the revenues created by the link-tax. The provision does not provide further details and therefore it is expected to be applied pursuant to the willingness of publishers. This is a bad news for journalists who maybe had dreamed to be better paid.

[Another article on the upload filters to follow] 

 

 

The status of the European copyright reform and the impact on publishers

comebacksnoopy-pg24-highres

UPDATE 5 February 2019: Germans and French delegations found a kind of agreement reported in a new compromises proposal by the Romanian presidency, whereby the exemption in favor of SMEs is eliminated while a soft-regime is granted to micro-entreprises (with a turnover less of 10 million Euro).  The latest draft of art. 13 is quite messy and will make very difficult life for videosharing platforms, unless they install overblocking filters. Censorship seems to be the most convenient and practical solution for operators wishing to avoid legal issues with rightsholders.

The European copyright reform is currently delayed because of disagreements amongst Member States about some crucial aspects of the new legislative text. The Rumanian presidency failed in proposing an acceptable compromise and its mandate request, which would have permitted to the Council to rapidly close the negotiations with Parliament and Commission, was rejected on January 18, 2019.

The apparent stallo between France and Germany

Apparently, the main stallo situation is about a potential exemption clause in art. 13 (the upload filters provision) for small enterprises: Germany defends the exemption, while France opposes. This is why Germany and France are reported to talk together and bilaterally, in order to find an agreement and speed up the end of the legislative process. The timing is crucial: the stakeholders supporting the reform (commercial broadcasters as well as traditional media and publishers) fear that the new European Parliament (to be elected in May 2019) may be less favourable to their interests, therefore they hope that everything should be finalised within the current legislature. To be fully sure to complete the legislative process without interferences, the plenary session of the European Parliament should approve the copyright reform by March 14, 2019.

The disagreement about the “small enterprise exemption for upload filters” seems to be a minimal question with respect to the ambit of application of the entire copyright reform. Therefore, it is more reasonable to believe that France and Germany are negotiating, with the support of the Commission, in order to define all pending details and getting a final agreement. The rumours about the matter to be escalating to Macron and Merkel are likely instrumental to this effect: to submit to the others a plan which is final and  not negotiable any longer. This byzanthine behaviour may be disliked by other Member States, however this is the way things often happen in Brussels. The weight of a joint French-German position may be so relevant to overcome whatever blocking minority in the Council.

What the impact for publishers?

If the above is true, and provided that legislators act rapidly, the copyright reforms may be approved before the “Copyright Barbarians” will conquest the next parliament. What could be the possible impact for the publishing industry?

Despite official declarations, most stakeholders (including publishers) are skeptical about the concrete outcome of this reform, whether or not it could provide a (at least monetary) solution for the crises of traditional media. At the current situation, it seems more a matter of principle and a “contentino”, namely a cake for a kind.

In this respect, on should consider whether the ancillary right provided by art. 11 of the new Directive could be waived by publishers (as it happened in Germany with the 2013 local legislation) or not (as it happened, instead, in 2014 in Spain). The European proposal has been always ambiguous to this point, although the current Council’s text seems to favour the German solution: “Nothing in this Directive should be interpreted as preventing holders of exclusive rights under Union copyright law from authorising the use of their works or other subject-matter for free, including through free licences, when they consider it appropriate” (recital 43b). To undermine this result, the Parliament proposed a different wording: “the listing in a search engine should not be considered as fair and proportionate remuneration”.

The German scenario (i.e.: publishers may waive the ancillary copyright)

In case of a “German scenario” publishers may waive the compensation from Google as well as other news aggregators (and, in general, from whoever should pay the fee because displaying an article or a simple snippet, entirely or via a hyperlink). It is well-known that small and innovative online publishers are against this reform and therefore will be more than happy to make the waive. But what would happen with the big publishers which have been supporting this reform? When the German copyright law was enacted, most publishers accepted the waive with Google but then some big of them filed a recourse with the antitrust authority for abuse of dominant position. Would this happen also with the European legislation and with which chances? Google is super-dominant in many markets, but not as Google News in the market of news aggregations, since this market is evolving and other platforms and communicating tools, including Facebook and Whatsapp, play an important role. To better attack Google and find it dominant (and potentially abusive), one should address the entire online advertising market (where also Facebook operates, however). This move, however, would weaken the competition case for the ancillary copyright, which is much more specific than online advertising. A competition assessments is always based on valid economics, not on lobbying declarations, and therefore there might be the risk that a competition authority states that a non-dominant Google News has a legitimate interests in offering the news aggregation service for free because it gains from it much less than what the publishers do. This is confirmed by a consistent perception that publishers need news aggregation and digital sharing irrespective from a potential remuneration via an ancillary copyright.

The Spanish scenario (i.e.: publishers cannot waive)

In case of Spanish scenario, publishers would not be entitled to waive their rights vis-à-vis Google and whoever. This is a complex scenario because a mandatory licensing system would be complicated to set-up and, in any case, the big fishes, Google in primis, still would have the choice between the nuclear option (closing the business limited to news aggregation) or grant the minimum. Fact is, it is doubtful whether the economics of the online news market may compensate publishers for what they have lost during the years. As mentioned by Quintarelli, the margin of Facebook in Europe per user is 1,3 Euro/month (with ARPU at 2,4 Euro), while for Twitter is only 10 Eurocent/month/. Margins may be higher for Google but, as mentioned above, the case is about Google News, not the entire’s Google business. In other words, there is no decent pie to share amongst the hundreds European traditional publishers.

The Spanish scenario may be further complicated by third parties setting the fee on the basis of various criteria, with subsequent appeals by opposing operators. The practical enforcement of the compensation will be delegated to the national authorities and the directive does not say very much about. The legal uncertainty resulting from that may likely undermine any potential gains.

The impact for small publishers and users

The above uncertain scenarios will be mostly problematic for small and innovative online publishers providing quality and local content. For them the Spanish scenario would be chaotic since they would not be able to set up decent licensing system with the entire market because they do not have sufficient resources to do it. Such operators will certainly be in favour of the German scenario (i.e.: the ancillary copyright may be waived). By contrast, poor informative or low-quality publishers may be interested with the Spanish system because their business is based on sharing whatever content and rapidly, from fake news to kittens.

The uncertainty will be detrimental for the Internet in general, in particular because the foreseen complex regulation of hyperlinks and snippets, with various conditions, carve-out and exemptions attached. To make an example, the text agreed sofar by the Council (recital 34) states that: “The rights granted to the publishers of press publications should not ….. extend to the mere facts reported in the press publications”. The need to rule this clear, self-evident principle, which is fundamental for the freedom of speech, is an explanatory evidence of potential disturbing consequences deriving from the regulation of hyperlinks and attached statements (the so-called snippets).

The hyperlinks 

The current drafts of  art. 11 (the so-called link-tax) keep uncertain the status for the hyperlinks. Article 11 (press publishers’ rights). The text of the Romanian Presidency insists with the quantitative criterion to exclude snippets, while other delegation would prefer a qualitative criterion. The application of a quantitative criterio may look like simpler in theory, however it will end up with arbitrary results: an hyperlink will be subject to a fee depending of the number of words attached to it. In addition, there is no clear carve-out for individual users (bloggers and so on) and micro-enterprises. An appropriate exemption is foreseen by the Parliament:

“the rights ……. shall not prevent legitimate private and non-commercial use of press publications by individual users”

however its practical implementation is doubtful due to uncertain border between personal commercial use in the Internet, because of adverttising, terms and conditions of blog services as well as professional interests mixed with private uses.

***

What next? Germans and French should soon let us know what they have decided under the support of the Commission. We understand that next COREPER I meeting is scheduled on February 8, while the Trilogue could take place on February 11. The last plenary session of the Parliament is scheduled in April 2019 but, for technical reasons, it would be advisable for the copyright legislator to close the file in March 2019.

 

 

The increasing uncertainty about mobile consolidation in Europe

 

20008557-two-traditional-dutch-pastry-called-tompouce-with-flags-over-white-background

 

Mobile consolidation within domestic markets has always been in the European telcos’ wish list. However, in the last 5 years this ambition hardly clashed with the European Commission’s approach, in the persons of Margareth Vestager, heading the Competition directorate (“DG COMP”) of the Commission, who treated the subject with severity during her mandate. Vestager opened the doors to panEuropean mobile mergers, but showed to be very strict towards domestic mergers (unlike her predecessors Almunia, who cleared 3 domestic mergers, such as in Germany, Austria and Ireland). To achieve this scope, Vestager made national mobile mergers (unlike transnational) subject to a remedy of divestment of spectrum and other mobile resources, in favor of a new entrant mobile operator, so as to maintain at least 4 network operators in the national market and avoid the reduction to only 3. This “rule” was consistently applied in important merger operations in Denmark, the United Kingdom and Italy, despite mobile operators claiming (in vain) that consolidation was necessary to support investments, especially for 5G.  DG COMP offices never bought this argument, and they may have some reason, considering the high sums paid by mobile operators in 5G auctions, irrespective of consolidation.  However, this approach of the Commission may also be criticized because it supports competition merely on pricing arguments, while the market needs to move towards more innovation and diversification.

But a few weeks ago, on 27 November 2018, the European Commission authorized, with some surprise, a domestic mobile merger happening in the Netherlands: it was about the purchase of the small Tele2 (5% of market share) by the third Dutch operator  T-Online (20% of market share), resulting in the reduction of mobile network operators in the Netherlands from 4 to 3. One wonders if this decision constitutes a precedent for starting a new approach in the merger policy of the European Commission, or whether it is an accident. At the moment this is not known, because European Commission’s officials are reported to exclude a revirement from the merger practice developed so far, while the Dutch case should be regarded as the result of exceptional circumstances. This approach, however, is misleading because the Commission did not provide any convincing arguments that the Dutch market conditions were so different from other cases dealt by the Vestager’s offices with much more severity. Therefore, this new case, rather than providing new guidelines for the future, only provokes uncertainties.

The merger between Tele2 and T-Online in the Netherlands led to the creation of a JV owning a consolidated market share of 25%, with the loss of the smallest and most competitive operator in the market (Tele2). One could believe that this scenario would strengthen competition, as the incumbents KPN and Vodafone / Ziggo would have now to face a stronger, consolidated third competitor (and no one else). However, the Commission’s competition practice developed so far (at least before this case) sounds different: the presence of a small group of operators (only 3) with similar market shares (between 25 and 35%) does not clearly help competition, while it may provoke the opposite effect, in particular the alignment of commercial practices – a normal behavior in small oligopolies. The traditional mantra of the Commission has normally been that competition is surely created by a Maverick operator, that is to say a small, hungry challenger who needs to be aggressive (with low prices but not only) to attract new customers and use efficiently its network (which otherwise would be empty). In the Dutch case the classic Maverick operator was Tele2, which with only 5% of the market had no chance but being very aggressive. However, thanks to the authorization of the Commission, Tele2 will now disappear, while it is not clear which incentive should bring the new JV to continue to be so aggressive, since with a consolidated market share of 25% it could find more convenient to align its pricing policy with the rest of the Dutch market. In previous cases, such as Italy and the UK, the Commission’s strict approach had led to the imposition of the entry of a Maverick operator, in these cases the French Iliad, which then actually entered Italy (but not in the UK). The entry of Iliad in Italy has caused a dramatic fall of domestic prices. One would correctly question why the entry of Iliad (or whatever low-cost operator) was imposed for Italy, and not for the Netherlands. 

The reason why the Maverick operator’s rule was disregarded by DG COMP in the Netherlands is unknown. The Dutch Tele2 was financially weak and poorly developed, it only had the 4G network and for the rest depended on national roaming agreements. It was a kind a super-MVNO, like Iliad in Italy. However, being small and week has never been an argument to decrease competition. By contrast, as previously said, Tele2 was the only operator in the Netherlands in real need to be aggressive to acquire new customers and market shares. Fact is, it was the only one to have offered an unlimited data plan. This type of competition will now disappear, although the new JV has offered to continue to maintain this aggressive offer: but it is unilateral commitment, not a remedy, which could be withdrawn any time.

No surprise that, at the news of the T-Online/Tele2 merger’s authorization, KPN’s shares jumped upwards, demonstrating that the disappearance of the Maverick operator in the Dutch market will weak competition and allow prices to rise.

If the European Commission had argued and explained that the time had come for a change of direction, that’s is to say: “welcome domestic mobile mergers”, then everyone would have understood. There are many reasons to support the idea that domestic mobile mergers make more sense than cross-borders ones. Fact is, the latest European regulatory developments suggest that the mobile market is and will remain marked by national borders: spectrum policy remains firmly in the hands of national authorities; the new European code strengthens the powers of national authorities to decide national cases on the basis of local specifications; wholesale caps roaming are well above retail domestic prices, thus preventing permanent roaming. In other words, European regulation has done very little to support a business case for pan-europeans mobile mergers, unlike domestic ones, therefore one should wonder why this scope should be boosted by a single leg of the European Commission (DG COMP), while the other legs are rowing against.

On the contrary, the DG COMP has made a striking exception to the merger practice implemented so far, but at the same time denying that something may have changed. In this way, the doubt remains that the Dutch case is a more political* rather than a market decision, and that there are no clear rules for the future. Belgian authorities are planning to open the market to a 4th mobile operators, while in other countries (France, Spain) there are reflections about consolidating down to 3. What lesson to be learned from this Dutch case? Boh.

*I would not be surprised to learn, one day, that this decision may have been taken at pure political level, with some senior officials dissenting or even opposing it. A similar situation likely happened during the Almunia’s term (2009-2014) when other 4 to 3 mobile mergers were authorized, despite the likely contrary opinion of the offices. However, there are no official evidence for that, since the eventual dissenting opinion of the offices cannot be reported in official minutes or drafts. We have to stay with doubts and suspects, made more inconvenient by the fact that the beneficiary of this strange decision is the German incumbent Deutsche Telekom.