Comme suite aux dernières mesures de confinement annoncées en Belgique, l’arrêté dit “de confinement” le plus récent, daté du 28 octobre 2020 et portant des mesures d’urgence pour limiter la propagation du coronavirus Covid-19, a été amendé le 1er novembre par un arrêté publié au Moniteur belge le jour-même.
Simont Braun a réalisé une version coordonnée de cet arrêté, celle-ci est disponible ici.
L’arrêté modificatif du 1er novembre renouvelle également l’annexe (p. 14) ; il est disponible ici en version complète bilingue français-néerlandais.
Nos équipes se tiennent par ailleurs à votre disposition pour toute question ou demande d’assistance suite à ces nouvelles mesures. Toutes leurs coordonnées sont accessibles ici.
Summary | This article deals with the private international law aspects in the Belgian Google case. With its decision of 14 July 2020, the Belgian Data Protection Authority confirms it is competent to hear a complaint filed against Google’s Belgian subsidiary, Google Belgium SA, even though the latter does not determine the purposes and means for the processing – which is determined solely by Google LLC, the mother company located in California. The reasoning of the Authority is three-ponged. First, it establishes that European data protection rules apply to the processing activity in question by applying the case law of the European Court of Justice. This results in the Authority having jurisdiction over the alleged infringement, conditional on the fact that the “one-stop-mechanism” would not be applicable. Then, the Authority establishes that the “one-stop-mechanism” does not apply and that there is thus no lead supervisory authority as the company responsible for the processing activity in question (i.e. Google LLC) is not established in the European Union. Lastly, it establishes that the complainant could bring a complaint only against Google Belgium SA because of the inextricable link between Google Belgium SA and its mother company Google LLC, the ambiguity created by Google themselves and the need for effective recourse for European data subjects. An appeal against the decision is currently pending before the Brussels Court of Appeal, section Market Court. A decision definitely worth keeping an eye out for.
Old, but not forgotten. On 14 July 2020, the Belgian Data Protection Authority (Autorité de protection des données / Gegevensbeschermingsautoriteit, hereinafter “BDPA” or the “Authority”) published a decision in a case involving Google following a failure to erase personal data as requested by a data subject (you can read the decision here). Google was found to breach multiple provisions of the General Data Protection Regulation (hereinafter “GDPR”). After an extensive review of the facts and interests at stake, the BDPA imposed a fine of 600k. Not surprisingly, the decision of the BDPA received widespread media attention. The fine imposed on Google is the highest fine to date.
The decision comprises an interesting aspect of private international law. Transnational companies pose multiple problems for national authorities. One of them is the complexity to determine where the processing takes place. The division of tasks within those companies is not always clear. Those questions were precisely what the Authority had to address in the case at hand. Before establishing a breach and being able to impose a fine, the BDPA first had to address whether it had jurisdiction as to such a breach under the GDPR. This is not new. The debate takes place in the bigger context of European countries struggling to ensure proper application of their national rules against transnational companies. It provides us with the opportunity to review the rules of jurisdiction and their functioning in the context of data protection.
The analysis of the BDPA is without a doubt of keen interest to many data protection lawyers, data protection officers, academics and companies, not only in Belgium but across the world.
After a short description of the case at hand (I), we will remind the applicable rules of jurisdiction and applicable law and its evolution in the context of data protection (II). Finally, we will describe how the BDPA applied them in its decision (III)
The facts leading to the case are quite simple and resemble a common scenario. When looking for information about the complainant on Google – by using his name and surname -, the search results referenced websites revealing personal information about the complainant that he wished others would not see when “googling” his name. The search results were deemed harmful to his honour and reputation.
More concretely, the case concerned a Belgian citizen, heading a big company at the time of the dispute, who had previously been in charge of several public positions and was a member of a Belgian political party, which a series of google search results referred to. In addition to that, some search results referred to articles that described how a complaint for harassment had been directed against him, which had been dismissed. Those websites were all referenced on Google.
The complainant sent Google a request to dereference a number of websites. He used the online form made available by the Google search engine created specifically for people to exercise their “right to be forgotten”, which is directly managed by Google LLC, the parent company located in California.
After examining the request, Google refused for various reasons, such as pages that were inaccessible, or which did not meet Google’s criteria for removal, but also on the ground of the public’s right to information, with considerations for the fact that the complainant is a public figure.
Following Google’s refusal, the complainant filed a complaint against Google’s Belgian subsidiary, Google Belgium SA, with the BDPA, aimed at obtaining the effective dereferencing of the websites at stake.
Preliminary central points of discussion during the proceedings were whether or not the complaint should have been directed against Google Belgium SA, Google LLC or Google Ireland Ltd and whether or not the BDPA had jurisdiction with regard to those parties.
Rules of jurisdiction and applicable law
Although the question of jurisdiction of Member States’ courts and authorities is not new and many issues had already been raised and solved under the regime of Directive 95/46/EC (see more in detail below), it is the first time the question of the BDPA’s jurisdiction is dealt with in Belgium.
Under the directive it was not mandatory for the Member States to grant corrective powers to their data protection authorities, such decision being left at their own appreciation. In Belgium, no such powers had been recognised to the Authority under the Privacy Act of 8 June 1992. Its role at the time was mainly advisory. One exception was the ability of the President of the Authority to file a suit with civil courts in Belgium.
Such power had been used to launch a case against Facebook Belgium BVBA, Facebook Ireland Ltd and Facebook Inc.. At the time, the Brussels Court of Appeal had dismissed it on grounds of lack of jurisdiction of the Belgian courts. Facing the absence of clear rules of jurisdiction, the Brussels Court of Appeal had applied the general regime of private international law to consider that:
The action of the Authority was of a public nature, which excluded the application of both the Brussels Ibis Regulation and of the Belgian Code of Private International Law;
Such exclusion still allowed the Court to have jurisdiction as to the complaints directed towards Facebook Belgium BVBA when it comes to its own behaviour since it was a purely internal situation. This, however, could only lead to sentencing actions which the Belgian company was directly and personally responsible for;
Also, the Court found no ground of jurisdiction as to the case filed against Facebook Ireland Ltd and Facebook Inc. and, therefore, declared itself incompetent.
This approach, combined with the lack of power of the Authority to impose administrative fines, actually led to a lack of effective recourse in Belgium against breaches committed by transnational companies.
The GDPR brought forward an improvement in this regard:
First, it gave corrective powers to all national data protection authorities of the different Members States, effectively offering an additional recourse to data subjects;
It drew clear rules of jurisdiction and applicable law, building upon earlier case law of the European Court of Justice (hereinafter “CJEU”) under Directive 95/46/EC; and
It created a general competence for “lead supervisory authority” when it comes to cross-border processing carried out by data controllers or processors whose main establishment is located in their territory.
Under the GDPR, the European privacy rules are applicable to “processing of personal data in the context of the activities of an establishment of a controller or a processor in the European Union, regardless of whether the processing takes place in the European Union or not” (art. 3.1 GDPR). They are also applicable to “processing of personal data of data subjects who are in the Union by a controller or processor not established in the Union, where the processing activities are related to: (a) the offering of goods or services, irrespective of whether a payment of the data subject is required, to such data subjects in the European Union; or (b) the monitoring of their behaviour as far as their behaviour takes place within the European Union” (art. 3.2 GDPR).
Article 55 GDPR provides that “each supervisory authority shall be competent for the performance of the tasks assigned to and the exercise of the powers conferred on it in accordance with this Regulation on the territory of its own Member State”. Recital 122 of the GDPR clarifies that this should cover in particular “processing affecting data subjects on its territory or processing carried out by a controller or processor not established in the Union when targeting data subjects residing on its territory”.
These provisions result from previous case law of the CJEU under Directive 95/46/EC. The first and most well-known case rendered was the Google Spain judgment (CJEU, 13 May 2020, C-131/12, Google Spain et Google Inc. v. Agencia Española de Protección de Datos (AEPD) et Mario Costeja González). In this decision, the Court considered that the promotion of the processing activities of Google Inc by its Spanish subsidiary (offering advertising space to make the search engine profitable) was sufficient for considering the latter as an establishment in the sense of Directive 95/46/EC. Consequently, in this case, Spanish law was applicable to the processing undertaken by Google.
Later, in its Weltimmo judgement, the CJEU stated that any national law on the protection of personal data applies where the data controller exercises, “through stable arrangements in the territory of that Member State, a real and effective activity — even a minimal one — in the context of which that processing is carried out” (CJEU, 1 October 2015, C-230/14, Weltimmo s. r. o. v. Nemzeti Adatvédelmi és Információszabadság Hatóság). Where their national law is applicable, data protection authorities may exercise effective powers of intervention (e.g. impose penalties) on the data controller or processor within the territory of their own Member State.
The combined effect of those cases strengthened the power of data protection authorities to enforce European data protection law on transnational companies, provided that the transnational companies exercised part of their activity through an establishment in that territory. This would apply even if the local entity’s activities merely consists in advertising the services of its parent company (the data controller). It is, therefore, not required that the said establishment actively participates in the processing of personal data, as later confirmed in the Wirtschaftsakademie judgment (CJEU, 5 June 2018, C-2010/16, Unabhängiges Landeszentrum für Datenschutz Schleswig-Holstein v. Wirtschaftsakademie Schleswig-Holstein GmbH).
It is the application of those rules that the BDPA had to consider for the first time in the present decision.
The analysis of the BDPA
As explained above, in the case at hand, the BDPA had to determine whether or not it had jurisdiction with regard to the refusal of Google to dereference the websites at stake.
First, the BDPA examined whether or not the GDPR was to be applied. Territoriality is an important principle of the GDPR. The territorial competence stems from the principle in international public law where a State only has competence to enforce the law on its own territory. On that point, when interpreting article 3 GDPR, the BDPA observed a legal gap. The law does not address the situation where a data controller having an establishment in the European Union does not process personal data as part of the activities of that establishment. Applying the case law of the CJEU, the BDPA concluded that the GDPR was applicable because (i) the complainant resided in the European Union, and that (ii) otherwise, if the GDPR did not apply, no adequate, nor full protection could be offered to data subjects.
The next step consisted in determining whether the “one-stop shop” mechanism of article 56 GDPR applied. In other words, whether the BDPA lacked competence because another data protection authority enjoyed a prevailing jurisdiction as a lead supervisory authority. Google argued that this was the case, since the company had chosen its main establishment in Ireland through Google Ireland Ltd. Going through a detailed analysis of the division of tasks between the various companies (Google Inc LLC, Google Ireland Ltd, Google Belgium SA) the BDPA concluded that Google Ireland was not responsible for the processing at stake, i.e. (de)referencing results on the Google search engine. The responsibility for the functioning of the Google search engine and its three phases, namely exploration, indexation and selection of results, falls solely with Google LLC. Google Ireland only processes data for modifying the search results based on the search history of users. It thus concerned a different processing activity than the one in dispute, for which Google Ireland Ltd could not be seen as the main establishment in the sense of article 4.16 GDPR. This line of reasoning has been confirmed by the Council of State in a parallel case against Google in France by a decision of 19 June 2020. The Council of State held that the “one-stop shop” mechanism was not applicable. It confirmed that the French data protection authority (“CNIL”) was competent to impose a €50 million sanction on Google LLC (you can read the decision here).
Finally, the BDPA had to consider whether or not it was competent to hear the complaint filed specifically against Google Belgium SA and not against Google Inc or Google Ireland Ltd. For this part, the BDPA relied on the Google/CNIL judgement of the CJEU (CJEU 24 September 2019, C-507/17, Google LLC v. Commission Nationale de l’Informatique et des Libertés (CNIL)) to allow the complaint by virtue of the inextricable link between Google LLC and Google Belgium SA, and the requirement of adequate and complete protection of data subjects. It was not disputed that Google Belgium SA constituted a “stable establishment” in Belgium. It was also established that the processing was indeed carried out in the context of the activities of the Belgian establishment. The BDPA highlights the international reach and the ambiguity created by Google itself which made it difficult to clearly differentiate the responsibilities of the different entities within the Google group. Therefore, even though Google Belgium SA does not determine the purposes and means for the processing – which is determined solely by Google LLC – the inextricable link and the need for effective recourses that had been highlighted in the Wirtschaftsakademie judgement justified that the complainant directed his complaint to Google Belgium SA alone. In the BDPA’s view, it is of little importance whether the processing of the data is actually performed outside the European Union by Google LLC employees.
Particular in the reasoning of the BDPA, is that “as its activities are inextricably linked to those of Google LLC, the Belgian subsidiary – given the role it plays and itself describes – can be treated in the same way as a data controller for processing conducted within the context of the operation of Google’s search engine and responses to delisting requests in Belgium”. In these circumstances, Google Belgium SA is responsible for ensuring the compliance of the GDPR in Belgium.
Other aspects of the decision
The decision also deals with other interesting aspects such as the balance of interests between the “right to information” and the “right to be forgotten”, the territorial reach of the dereferencing request (cf. Google/CNIL judgement) and the criteria for determining the height of the fine.
The decision was published on the website of the BDPA because of “the importance of transparency in decision-making process and the decisions of the Litigation Chamber, and in light of the scope of this decision, which concerns a very large number of data subjects – i.e. all Belgian residents, and by extension all residents of the EEA – who might be listed by Google’s search engine in search queries using their first and last names as keywords”. In the case at hand, the BDPA has decided not to redact Google’s identifying information. This is exceptional. The BDPA is of the opinion “that these redactions are necessary to the pursuit of plaintiff’s objective, which is to be delisted by Google”. The argumentation of Google Belgium SA that the publication of the decision would be counterproductive and would stigmatise Google was rejected. In the reasoning of the BDPA “it is relevant to give this decision sufficient publicity to raise awareness among internet users of their rights under the GDPR”.
The decision, the first one by the BDPA addressing its international jurisdiction since the GDPR entered into force, gives much food for thought. The BDPA confirms it is competent to hear a complaint filed against Google’s Belgian subsidiary, Google Belgium SA, even though the latter does not determine the purposes and means for the processing – which is determined solely by Google LLC, the mother company located in California -.
In its reasoning, the BDPA considers all the existing case law and incorporates it in the recent legal development of data protection law. The BDPA comes up with a practical solution combining a strict legal formalism and the needed considerations for the right to an effective recourse recognised by the GDPR. For David Stevens, the president of the BDPA, this case is of great importance:
“This decision is not only important for our Belgian citizens, it also demonstrates our ambition to better protect online privacy together with our fellow European regulators. Concrete actions against such global players are therefore required. In this way, we want to actively contribute to a true data protection culture on a European level as well.” (you can read the article here)
The decision of the BDPA has the potential to significantly impact the way in which global organisations should be thinking about their data protection strategy in Europe.
It remains to be seen if this line of reasoning will be confirmed by the Brussels Court of Appeal, section Market Court. In appeal, it is highly likely that a prejudicial question will be asked to the CJEU regarding the interpretation of the GDPR. In the parallel case in France, Google requested the Council of State – but was denied – that the following question would be asked:
“Can a controller established in a country outside the European Union with several establishments in the European Union and a designated European registered office in the territory of a Member State have a “main establishment” within the meaning of Article 4(16) of the GDPR in that Member State in the event that decisions on the purposes and means of processing are taken in that third country?”
We look forward to seeing more development on this matter.
With the contribution of Viktor Francq, summer intern 2020 at Simont Braun
 Brussel, 8 mei 2018, R.D.C.-T.B.H., 2020, nr. 1, p. 75.
 Regulation (EU) No 1215/2012 of the European Parliament and of the Council of 12 December 2012 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters, Pb.L. 20 December 2012, issue 351, 1.
We have been hearing about autonomous vehicles (“AVs”) for quite a while. However, as time goes by, the use of AVs is getting closer to becoming a reality. In various countries, car manufacturers are now carrying out tests on open or closed parts of public roads. Just like engineers are facing technical challenges in this field, lawyers have to deal with questions around liability, cybersecurity, data protection, insurance, intellectual property, etc. In this framework, news about AVs accidents are getting less and less uncommon, and they lead to tricky liability questions, such as: can a car user be held liable if he or she had no direct control over the wheel?
Such questions might need a clear answer in the near future and the whole AV subject at large is gaining track. Recently, a Belgian politician launched the idea to make Belgium the first country to welcome AVs. In parallel, the European Commission issued in February 2020 a White Paper on Artificial Intelligence, referring to liabilities incurred by AI, and by extension by AVs.
What do you mean, “autonomous” vehicle (“AV”)?
AVs’ autonomy varies in degree. Subject to a specific European scale of autonomy measurement, authors generally refer to the Society of Automotive Engineering (“SAE”)’s scale. SAE is an international organisation based in US. It brings together engineers, CEOs, researchers, professors, and students who share ideas on automotive engineering. It established six levels of autonomy, ranging from a complete human intervention (level 0) to a total absence of human action (level 5).
To date, the most advanced AVs on the market are of level 3. In that case, the vehicle can handle itself all aspects of the driving tasks within a certain set of circumstances. In some instances, the human driver must be ready to take back control of the vehicle when the AV so requires, e.g. during a traffic jam. According to car manufacturers’ (enthusiastic) declarations, the next levels of autonomy should reach our market in the coming months or years.
How does Belgian law address liability issues in the case of AV accident?
At this stage, Belgian law does not have any specific liability regime addressing the risks generated by AVs.
A victim of an AV should thus find its way through the existing and non-specific Belgian liability regimes. In short, four liability regimes can be identified as potential legal basis for a claim.
1. The fault-based liability regime (Article 1382 of the Civil Code)
This requires the victim to prove three things: a wrongful behaviour from the AV user, a damage and a causal relationship between the behaviour and the damage.
The tricky question here is whether a driver can be found guilty of a wrongful behaviour if he or she was letting the AV operating itself at the time of the accident?
2. The strict liability regime for the use of defective things (Article 1384 of the civil Code)
The victim should prove its damage and a causal relationship, just like in the previous fault-based liability regime. However, in this case, the victim can limit itself to proving a malfunctioning in the AV, rather than proving a wrongful behaviour of the AV user.
This solves the question raised by the previous regime but raises a new one, e.g. how do you prove a malfunctioning when the accident is the result of an algorithmic decision, which in turn could potentially be the outcome of machine learning?
3. The product liability regime (law of 25 February 1991 on product liability)
This regime is similar to the second liability regime as the victim must prove its damage, a malfunctioning of the AV and a causal relationship between these two aspects. However, in this case, the victim will seek compensation against the manufacturer, not the driver.
Furthermore, the compensable damage will be limited to personal injuries (including moral damage) and, subject to certain conditions, damages to property. Other types of damage (e.g. loss of opportunity) are not recoverable under this specific liability regime.
4. The insurance liability (Article 29bis of the law of 28 November 1989 on compulsory motor vehicles liability insurance).
Once a vehicle is involved in an accident, “weak users” (e.g. pedestrians and cyclists) may obtain compensation from the car insurer for the damages resulting from this accident.
In this case too, recoverable types of damage are limited. They only cover personal injuries or death, as well as damage to clothing (quite oddly).
Vehicles are becoming more and more autonomous. Level 5 (fully autonomous) AVs are still prohibited at this stage but their eventual arrival on the market will generate liability questions for which our legal system is not well equipped.
They will change the fault-based paradigm and we anticipate that a dedicated liability regime (e.g. based on the mere use of the AV) will be needed to complement our existing legal framework. We at Simont Braun will closely follow up on this and keep you posted.
For any question, please contact the authors:
Ce mardi 17 mars, la loi insérant un nouveau livre 3 « les biens » dans le Code civil a été publiée au Moniteur belge.
La réforme maintient l’essentiel du régime applicable tout en le restructurant et en y insérant plusieurs lois spéciales telles que celles sur le droit de superficie et le droit d’emphytéose.
Lignes directrices de la réforme
Le remaniement ainsi opéré s’articule autour des quatre lignes directrices suivantes : (i) la modernisation et l’actualisation des termes utilisés (ii) une approche davantage simplifiée et intégrée, (iii) l’instrumentalisation et (iv) la flexibilisation du droit des biens.
Le livre 3 se décompose en sept titres dont le titre premier énonce un ensemble de dispositions générales constituant le futur ‘tronc commun’ du droit des biens s’appliquant sauf disposition contraire à tout droit réel. Le législateur a ainsi entendu uniformiser les règles relatives à la naissance, à l’extinction, à l’objet et aux effets des droits réels tout en maintenant certains régimes particuliers pouvant contenir des dérogations aux normes générales.
Ce titre premier détaille certains des principes directeurs de la réforme.
L’article 3.1, répondant au besoin de flexibilité contractuelle, énonce ainsi que les dispositions du livre 3 sont supplétives sauf s’il s’agit de définitions ou si la loi en dispose autrement. C’est notamment le cas des dispositions relatives à la copropriété forcée qui sont expressément qualifiées d’impératives.
L’article 3.3 maintient quant à lui le controversé numerus clausus en vertu duquel seul le législateur peut créer des droits réels. Ceux-ci sont limitativement énumérés. Il s’agit du droit de propriété, de la copropriété, des droits réels d’usage et des sûretés réelles. La catégorie des droits réels d’usage est quant à elle formée des servitudes, du droit d’usufruit, du droit d’emphytéose et du droit de superficie.
L’on épinglera encore l’article 3.4 qui consacre la règle selon laquelle, sauf exception, en cas de conflit entre des droits réels, le droit antérieur prévaut sur le droit postérieur.
La réforme consacre également différents régimes particuliers tels que la copropriété (titre 4), l’usufruit (titre 6) ou encore les relations de voisinage (titre 5) qui se voient désormais dotés d’un corps de règles centralisé et uniformisé.
Troubles de voisinage – intégration des principes jurisprudentiels et doctrinaux
Le titre 5 prévoit désormais légalement, en son article 3.101, la théorie jusqu’alors jurisprudentielle des troubles de voisinage selon laquelle tout titulaire d’un attribut du droit de propriété qui rompt l’équilibre établi entre fonds en infligeant à son voisin un trouble qui excède la mesure des inconvénients normaux du voisinage peut être condamné à l’une des mesures énoncées au paragraphe 2 de cette disposition.
Au-delà des sanctions, ce titre encadre précisément le champ d’application, l’appréciation du caractère excessif ou encore la prescription de l’action en matière de trouble de voisinage.
Intégration et modernisation des régimes d’emphytéose et de superficie
Parmi les évolutions notables, l’on soulignera encore l’insertion de deux titres distincts relatifs au droit d’emphytéose (titre 7) et au droit de superficie (titre 8) qui étaient jusqu’alors règlementés par les lois du 10 janvier 1824 modifiées en 2014. Ces deux titres sont supplétifs de volonté à l’exception des définitions et des dispositions qui en régissent la durée.
Le législateur a profité de la réforme pour clarifier et expliciter la durée, l’objet ou encore les droits et obligations de l’emphytéote et du superficiaire.
En ce qui concerne le droit d’emphytéose, il a ainsi précisé que son titulaire dispose de l’usage et la jouissance matériels et juridiques de l’immeuble sur lequel porte son droit. Il peut, dans l’exercice de ce droit, réaliser tous les ouvrages et plantations qu’il souhaite, et ce même s’il modifie de cette manière la destination de l’immeuble, pour autant toutefois qu’il n’en diminue pas la valeur. Pour ce faire, l’article 3.173 énonce que l’emphytéote doit faire toutes les réparations d’entretien et toutes les grosses réparations au sens de l’ article 3.153 régissant les obligations d’entretien de l’usufruitier.
Quant à son objet, le droit d’emphytéose peut porter indifféremment sur un fonds de terre bâti ou non ou sur un volume en sursol ou en sous-sol.
Enfin, la durée minimale du droit d’emphytéose, jusqu’alors fixée à 27 ans, passe désormais à 15 ans alors que sa durée maximale demeure de 99 ans.
Quant à elle, la durée maximale du droit de superficie est désormais alignée sur celle du droit d’emphytéose et ainsi étendue à 99 ans alors qu’elle est actuellement limitée à 50 ans, ce qui peut poser problème lorsque les dissociations de propriété sont envisagées pour une longue durée.
A titre exceptionnel, le droit de superficie peut être perpétuel. Tel est le cas lorsque le droit de superficie est constitué par le propriétaire du fonds soit à des fins de domanialité publique, soit pour permettre « la division en volumes d’un ensemble immobilier complexe et hétérogène comportant plusieurs volumes susceptibles d’usage autonome et divers qui ne présentent entre eux aucune partie commune » (article 3.180.). Cette seconde hypothèse suppose que coexistent sur un même fonds au moins deux volumes comprenant des ouvrages destinés à des affectations diverses susceptibles de gestion autonome et ne présentant aucune partie commune constitutive d’accessoire nécessaire (à titre d’exemple, les travaux préparatoires font référence à un bâtiment résidentiel, à une place publique ou à un centre commercial avec parking souterrain).
Portée de la réforme
En conclusion, la réforme des biens ne constitue pas une véritable révolution, de l’aveu même de ses deux architectes, les Professeurs Pascale Lecocq et Vincent Sagaert. La plupart des dispositions nouvelles entérinent en effet des évolutions jurisprudentielles et doctrinales déjà bien établies.
Il convient toutefois de ne pas en minimiser la portée.
Le futur livre 3 entrainera des répercussions aussi bien pour les professionnels de l’immobilier dans la manière dont ils structureront leurs opérations (avec notamment la modification des régimes de l’emphytéose et de la superficie permettant entre autres la création de volumes) que pour la vie quotidienne des particuliers (copropriété, voisinage ou usufruit).
Entrée en vigueur
La loi entre en vigueur le premier jour du dix-huitième mois qui suit celui de sa publication au Moniteur belge, soit le 1er septembre 2021.
Introduction and context
It is safe to say that on 30 January 2020, the Court of Justice of the European Union (hereafter the “CJEU”) handed down a landmark decision in EU IP and competition law. The CJEU writes a new chapter in the patent dispute settlements saga, and more particularly in the debate commonly known as “pay-for-delay”. To recall, in the context of the pharmaceutical sector, “pay-for-delay” is a term used for agreements between originator pharmaceutical companies and generic manufacturers where the latter agrees for a certain period not to enter the market with generic medicines in exchange for a value transfer.
The question arose whether there was an issue with competition law. It is particularly hard to answer when it takes place in the context of a settlement between the originator company and the generic manufacturer following a real dispute (i.e. patent dispute settlements). Ever since the European Commission launched its sector inquiry regarding patent dispute settlements that ended in 2009 (the full first report available here and the latest report of 2018 here), where the issue was brought to the surface, market players and practitioners have been eager to know the CJEU’s take on the matter.
There has been a belief that competition rules on restrictive agreements do not apply to settlement agreements between originator pharmaceutical companies and generic manufacturers who have not yet entered into the market as the companies are not in competition. The pharmaceutical context – and patent law specifically – would constitute an insurmountable barrier to enter the market. There is also the belief that, even if competition law would apply, agreements to settle ongoing court proceedings pursue a legitimate objective which is from the outset incompatible with the categorisation of an agreement as restricting competition by its object, since such settlement agreements have a public interest and are encouraged by the public authorities.
The question of whether a patent dispute settlement agreement would constitute a restrictive agreement by object or by effect is of great importance. When it is considered a restrictive agreement by object, the restrictive effect of the agreement on competition does not need to be proven for there to be an infringement of competition law, as opposed to when it is considered a restrictive agreement by effect.
By way of preliminary ruling, the CJEU, in line with the reasoning of AG Kokott (which you can read here), sheds light on the matter and inter alia (figuratively) bursts these two assumptions. You can read the judgement here.
By way of information, the present case takes place within the context of Lundbeck (case T-472/13) and Servier (T-691/14), currently pending before the CJEU, in which the European Commission found that agreements in settlement of patent disputes constituted infringements of article 101 of the Treaty on the Functioning of the EU (“TFEU”) and with regard to Servier, article 102 TFEU.
It should also be mentioned that the debate is not only taking place in Europe but also elsewhere in the world. For instance, in California (US), Assembly Bill No. 824 – commonly referred to as the “Pay for Delay” bill (here) – passed, which makes it unlawful for companies to settle patent infringement claims filed by generic manufacturers by providing “anything of value” in exchange for settlement. Any attempts to settle in such a fashion will be considered anti-competitive and open the company to civil litigation. The bill is aimed at lowering the cost of prescription medicines and fostering greater access to healthcare. It became effective on 1 January 2020. Not completely surprising considering that the Supreme Court of the United States had already found in 2013 that these settlements could violate antitrust laws (here).
The position of the CJEU
In the judgement of 30 January 2020, the CJEU responds to 10 questions (which the CJEU boiled down to 5) referred to by the Competition Appeal Tribunal (UK). The latter wonders, in essence, whether an agreement to settle a medicinal product patent dispute constitutes a restriction of competition by object or by effect and whether the conclusion of that agreement, possibly combined with entry into other agreements, constitute an abuse of dominance.
Before going into the decision, it is important to recall the facts briefly. It concerns an originator pharmaceutical company that is the holder of a manufacturing process patent for an active ingredient that is in the public domain, and generic manufacturers who have not yet entered but are preparing to enter the market with generics having the same active ingredient. The originator pharmaceutical company and the generic manufacturers were in dispute as to whether the process patent is valid and/or whether the generics infringe the patent.
Given that the originator pharmaceutical company only held a process patent, there was uncertainty of whether or not there was an infringement. Against that background, the originator pharmaceutical company and the generic manufacturers, even the ones with whom at that time no patent litigation was yet pending, concluded settlement agreements. The generic manufacturers undertook not to enter the market and not to pursue their actions challenging the validity of the patent for the duration of that agreement, in return for transfers of value.
In light of those circumstances, the CJEU responds very precisely by ruling that:
- the pharmaceutical company and the generic manufacturers are potential competitors provided that the latter have a firm intention and an inherent ability to enter the market, and do not meet barriers to entry that are insurmountable;
- the settlement agreement where the net gain has no other explanation than the commercial interest of the parties not to engage in competition on the merits, constitutes an agreement that has as its object the prevention, restriction or distortion of competition unless the settlement agreement is accompanied by proven pro-competitive effects capable of giving rise to a reasonable doubt that it causes a sufficient degree of harm to competition;
- the strategy of the dominant undertaking, which leads it to conclude settlement agreements, which have, at least, the effect of keeping temporarily outside the market potential competitors who manufacture generics, constitutes an abuse of dominant position, provided the strategy has exclusionary effects going beyond the specific anti-competitive effects of each of the settlement agreements.
What should you take away from the judgment?
The decision is without a doubt of significant interest to pharmaceutical company patent holders, generic manufacturers as well as competition and IP lawyers. It provides much food for thought for those presently involved in or contemplating settlement agreements.
In general, the decision teaches us that patent dispute settlements are not immune from competition law review. In particular, it will be important to keep in mind that the mere fact of a dispute between a pharmaceutical company patent holder and a generic manufacturer may be sufficient to demonstrate that they are potential competitors and thus any settlement between them involving a value transfer is likely to attract antitrust scrutiny.
In that regard, the mere existence of the patent, or the likelihood of success in litigation (for example, in proceedings about validity or whether a patent has been infringed) is irrelevant to whether the parties are potential competitors and whether an agreement can affect competition. The question is rather whether, notwithstanding the existence of the patent, the generic manufacturer has real and concrete possibilities of entering the market at the relevant time. To that end, the CJEU explicitly recognised the ‘at risk launch of a generic medicine’ as inherent to the pharmaceutical market.
Positively, the CJEU affirms that patent dispute settlements are not as such prohibited by competition law and neither is the transfer of value in that setting. However, there is a high probability that where agreements involve significant value transfers and the sole consideration given for that value transfer is the generic manufacturer refraining from market entry and from challenging the patent during an agreed period, there will be a finding of a ‘restrictive agreement by object’ prohibited under article 101 TFEU.
The attentive reader might have spotted that the CJEU, as it only responds to the facts presented by the referring judge, leaves certain issues open. For example, one might wonder how the analysis might change if it was not a process patent but a compound patent or whether the transfer of value is not made solely in return for market exit and no-challenge commitments, but reflects other considerations between the parties.
Another interesting issue to further develop is the consequences of pro-competitive effects on the finding of a ‘restrictive agreement by object’. Of great value is the explicit recognition by the CJEU that pro-competitive effects are not completely excluded from the debate of whether or not there is a ‘restrictive agreement by object’. What is certain is that the CJEU does not require a balancing test to be carried out or consider that the mere existence of pro-competitive effects would be sufficient to exclude the finding of a ‘restrictive agreement by object’. The pro-competitive effects should be considered with the other circumstances of the case to verify if it raises a reasonable doubt on the conclusion that the agreement, by its nature, would result in a sufficient degree of harm to competition and whether it should thus be categorised as a ‘restrictive agreement by object’.
We look forward to the CJEU further expanding on the above and its guidance on patent dispute settlements in general when it delivers its Servier and Lundbeck judgement.
Eric De Gryse and Christopher Dumont
You may always contact the authors should you have any questions:
email@example.com – +32 2 533 17 52
firstname.lastname@example.org – +32 2 533 17 58