A. Introduction
In 1984, European Court of Human Rights Justice Pettiti, in his concurring opinion in the decision of Malone v The United Kingdom, affirmed: “The danger threatening democratic societies in the years 1980-1990 stems from the temptation facing public authorities to see into the life of the citizens.” Footnote 1 This feared prophecy, unfortunately, was revealed to be accurate, and the 2013 Snowden revelations made it crystal clear: in “times of stress,” Footnote 2 characterized by terrorist attacks and transnational crimes, public authorities have increasingly resorted to surveillance systems as important tools to guarantee national and public security. Despite their potential to fight against terrorism and global crime, these instruments create serious intrusions into the private lives of millions of individuals; in particular, they endanger the right to data protection together with other rights and principles fundamental to the core of our democratic societies, such as freedom of expression and the principle of presumption of innocence. Unlike “targeted” surveillance, mass surveillance allows for the generalized and indiscriminate monitoring of every person with no need for a specific or reasonable suspicion and without an objective link connecting those subjected to control or a particular risk or crime. In recent years, the extensive employment of new technologies and big data enabling forms of mass surveillance for the sake of security has expanded already existent criticalities and added new issues related to both the increased ability to capture a vast amount of data and the capacity to automatically analyze them in an aggregated way, finding connections, determining social relations, habits, preferences, and movements; Footnote 3 in other words, the profiling of users. These sophisticated techniques add further complexities that make it even more urgent to determine a clear balance between general interests, such as national security, and safeguarding fundamental rights.
In the EU, this complex challenge is central to a heated debate involving civil society, courts, legislators, and law enforcement authorities, at national and supranational levels. The quest for a ban, or at least a more stringent set of rules governing the use of highly invasive surveillance measures has characterized the reactions of citizens and NGOs who are worried about the possible detrimental impact on fundamental rights and freedoms of preventive data and metadata control by public authorities – in particular, law enforcement and intelligence agencies.Footnote 4 In this context, civil society has played a decisive role. On the one hand, civil society groups triggered a vast judicial discussion on the necessity to determine proper and precise safeguards for mass surveillance techniques. Most of the landmark decisions of the Court of Justice of the European Union (CJEU), as well as national courts, were initiated by NGOs. They instigated a far-reaching momentum into this serious debate by pushing policymakers to (re-)think about the possibility of massively collecting, retaining, accessing, and processing private information and data for security purposes. On the other hand, civil society groups are nourishing a debate outside national and supranational courtrooms and parliaments on how to articulate the core principles of contemporary constitutionalism in the fight against global crime. Mass surveillance was merely a theoretical possibility when most constitutional charters were adopted. As this Article explains, it is rare for constitutional charters to directly limit mass surveillance. Civil society groups, through the adoption of non-legally binding declarations of digital rights, are fostering a conversation on how to ensure that traditional constitutional rights – such as the rule of law, due process, the right to privacy, and freedom of expression – are also guaranteed while deploying digital technology surveillance mechanisms for national or public security purposes.
The core argument of this Article is that these recent developments, comprehensively analyzed, denote a move towards a gradual, multilevel constitutionalization of mass surveillance. In this Article, we use this term to describe the translation and incorporation of already affirmed core constitutional values, particularly the principles of proportionality, legality, and the necessity of judicial oversight, in the specific context of state surveillance. Constitutionalization is used here as a lens to map the ongoing multilevel process of progressively defining constitutional limits and societal affordances of mass surveillance practices in the EU. Our Article is structured in three parts, focusing on the three main actors that so far triggered and shaped the constitutionalization of mass surveillance in the EU: civil society, the CJEU, and national legislators.
Section B reconstructs the role of civil society actors. We explain that surveillance practices have emerged thanks to a constitutional vacuum. Existing constitutional texts rarely address surveillance practices explicitly. This has allowed state authorities to potentiate and keep surveillance systems in place even in the aftermath of global scandals that made apparent the level to which these practices seriously violate core fundamental rights. Through an analysis of the initiatives and demands of civil society in the field of state surveillance, we show how this constitutional vacuum is no longer justified as it reflects the social reality of a time when mass surveillance was considered an unrealizable dystopia. Civil society organizations have played a crucial role in prompting a reaction to this normative void – not only by promoting strategic litigation contesting mass surveillance practices but also by explicitly advocating for a ban on mass surveillance through the adoption of non-legally binding declarations of digital rights.
Section C focuses on the constitutional input provided by the CJEU. Notwithstanding the specific demands from civil society actors, an analysis of the complex case law of the CJEU affecting state surveillance practices reveals that an absolute prohibition of mass surveillance practices does not appear to be a feasible and realistic solution. Conversely, the CJEU has put significant effort into actively defining the legal boundaries of mass surveillance practices in the EU. The constitutionalizing role played by the Court has thus consisted of constantly striving to find an equilibrium between Member State interests and fundamental rights, becoming both a defender of civil society demands and a pragmatic interpreter of the exigencies of public and national security authorities.
Section D examines the reticent – if not to say, at times ostensibly deaf – posture (still) adopted by national legislators in the field of surveillance practices, which is inexorably slowing down their process of constitutionalization. Civil society and judicial impulses have progressively impacted the normative choices available to national legislators. The carte blanche on surveillance powers accorded to national legislators in the aftermath of various terroristic attacks in Europe in the early 2000s is no longer available. In light of a growing body of delimitations developed by the CJEU, national legislators are now called on to pay more attention to finding pragmatic ways of striking a balance between national and public security and guaranteeing the fundamental rights at stake. Yet we conclude by denouncing how national legislators remain reluctant to incorporate the constitutional standards progressively developed by courts, thus observing the persisting ongoing nature of the multilevel process of constitutionalization of state mass surveillance practices.
B. Civil Society Demands: Invoking a Constitutional Ban on Mass Surveillance
I. In the Silence of Constitutions: The Normative Vacuum about Surveillance
Despite the “historical pedigree” Footnote 5 of the rights to data protection and privacy in Europe, which emerged as a counteraction to the atrocities committed by the authoritarian regimes of Western and Eastern European countries before and after WWII, an absolute prohibition of mass surveillance is not explicitly enshrined in any national constitution. This can be mentioned as one factor that has allowed mass surveillance practices to survive recent scandals and court cases. Data protection legislation and the constitutional right to data protection emerged in the EU as a normative response to the risks associated with increased automation, collection, and exchange of personal data in the public and private sectors. Footnote 6 Yet, despite the daunting potential of these technologies in the 1960s and the multiple contemporary lessons learned from authoritarian regimes in European history, surveillance is not explicitly addressed at the constitutional level.
The term “surveillance” is explicitly included in only two European national constitutions: the Swedish Constitution and the German Basic Law. The Instrument of Government, one of the four fundamental laws composing the Swedish Constitution, in Chapter 2, Article 6 reads:
Everyone shall be protected […] against examination of mail or other confidential correspondence, and against eavesdropping and the recording of telephone conversations or other confidential communications. In addition to what is laid down in paragraph one, everyone shall be protected in their relations with public institutions against significant invasions of their privacy, if these occur without their consent and involve the surveillance or systematic monitoring of the individual’s circumstances. Footnote 7
Paragraph 2 of this provision was added a decade ago after a recommendation from an ad hoc Commission on the Protection of Personal Privacy. Interestingly, the Swedish legislator added a provision that is almost redundant after protecting against the examination of all types of confidential communications, but that explicitly targets surveillance systems. Indeed, the norm of paragraph 2 explicitly refers to forms of “systematic monitoring” performed by “public institutions” without the individual’s consent that produce “significant invasions” of personal privacy. Yet paragraph 2 does not seem to ban mass surveillance. By reading Article 20, one understands that the explicit protection afforded by the Instrument of Government is the application of the principle of legality to potential restrictions on the rights enshrined in Article 6. In other words, the principle of secrecy of correspondence and the right to privacy are explicitly considered relative rights in Swedish constitutional law, thereby tolerating potential compressions that must be clearly defined by law, and not extemporaneously implemented by law enforcement authorities.
The German Basic Law adopts a similar approach. Surveillance is explicitly mentioned in the German constitution and is presented as a permissible practice subject to a series of guarantees. The term “surveillance” is used in Article 13, which enshrines the inviolability of the dwelling and refers to acoustic monitoring systems. Paragraphs 3 and 4 read:
(3) If particular facts justify the suspicion that any person has committed an especially serious crime specifically defined by a law, technical means of acoustical surveillance of any home in which the suspect is supposedly staying may be employed pursuant to judicial order for the purpose of prosecuting the offence, provided that alternative methods of investigating the matter would be disproportionately difficult or unproductive. The authorization shall be for a limited time. The order shall be issued by a panel composed of three judges. When time is of the essence, it may also be issued by a single judge.
(4) To avert acute dangers to public safety, especially dangers to life or to the public, technical means of surveillance of the home may be employed only pursuant to judicial order. When time is of the essence, such measures may also be ordered by other authorities designated by a law; a judicial decision shall subsequently be obtained without delay. Footnote 8
According to the Basic Law, surveillance is justified to investigate only serious crimes or dangers to public safety. A legal basis and judicial authorization are necessary to carry out surveillance practices. Surveillance is presented as a last resort practice, which can be used where no less invasive methods are available. Moreover, surveillance practices shall be limited in time. In sum, the German Constitution, in relation to acoustic surveillance of a dwelling, does not seem to avail a model of mass surveillance but that of targeted monitoring limited in time.
Interestingly, one can notice that the normative architecture of the German Basic Law follows a classical approach, dealing with the inviolability of the dwelling and the privacy of correspondence in two separate provisions. Article 10 indeed focuses on “correspondence, posts and telecommunications” and, interestingly, enshrines a different, lighter standard of protection in relation to surveillance practices affecting this portion of our private life. Paragraph 2 of this norm reads:
(2) Restrictions may be ordered only pursuant to a law. If the restriction serves to protect the free democratic basic order or the existence or security of the Federation or of a Land, the law may provide that the person affected shall not be informed of the restriction and that recourse to the courts shall be replaced by a review of the case by agencies and auxiliary agencies appointed by the legislature Footnote 9
The principle of legality is still the main guarantee the German Constitution introduces to protect the privacy of communications. However, in this case, the Basic Law does not refer to judicial authorization or time limits. Contrariwise, the scope of the justifications of a potential restriction is very broad: the protection of the “free democratic basic order” and the “existence or security of the Federation or of a Land” instead of specific serious crimes. Second, judicial oversight is expressly replaced by law enforcement or an intelligence agency review. Third, covert surveillance is explicitly authorized since the law may require that the persons affected by the ongoing monitoring are not to be notified. In sum, apart from the requirement to find a legal basis, Article 10 does not circumscribe the possibility of introducing mass surveillance practices.
This brief analysis of constitutional provisions reveals that, while protection of private life is generally enshrined in contemporary constitutions, no explicit provisions significantly limit the possibility of resorting to mass surveillance practices, even in constitutional texts that explicitly regulate surveillance. On the contrary, national constitutions generally include provisions that potentially restrict the principle of secrecy of correspondence, the right to privacy, and the right to data protection. None of these rights are presented as absolute. They all tolerate restrictions in the name of public and national security, thus de facto opening the door to the adoption of mass surveillance systems. In light of this consideration, we contend that constitutional provisions currently in force reflect the technological state of the art of a time when mass surveillance was still utopic and, therefore, do not afford adequate protection against the potential risks of these practices. However, in this Article, we posit that a progressive process of constitutionalization is currently underway in relation to state surveillance. A series of impulses are pushing towards developing constitutional principles, limiting the possibility of resorting to mass surveillance practices, and defining acceptable levels of compression of fundamental rights. In the following sections, we will analyze civil society and judicial impulses, investigating to what extent civil society groups and courts in the EU are normatively shaping constitutional guarantees against surveillance.
II. Multilevel Constitutionalization and the Driving Role of Civil Society
The constitutionalization process that is currently shaping surveillance systems is in reality a broader phenomenon. It encompasses all major changes and threats prompted by the digital revolution. It is plural, yet unitary in its ultimate objectives. Footnote 10 Recent scholarship analyzed the emergence of “digital constitutionalism” as an ideology informing and leading a multistakeholder movement. Footnote 11 One cannot identify a single constitutional framer, but several actors contributed at multiple levels to the debate on how to articulate constitutional norms that should preserve our fundamental rights in the digital ecosystem. Not only traditional, institutional actors, such as courts and legislators, but also new voices, often neglected from a legal point of view, such as civil society.Footnote 12 It is not a chaotic conversation but a process where each party stimulates and complements each other as the tesserae of a single mosaic. It is a multilevel process of constitutionalization that finds its unity in its aim to instill the core principles of contemporary constitutionalism in the context of the digital society.
Absent explicit, detailed regulation of surveillance at the level of national and supranational constitutions, civil society actors have played a driving role in the constitutionalization of mass surveillance. First, it is worth highlighting the important function that civil society played as a trigger of judicial cases. Indeed, the major decisions from national and European courts related to state mass surveillance practices, which we will analyze in more detail in the next section, were initiated by privacy activists, associations, and NGOs. Austrian privacy activist Max Schrems, Irish-based association Digital Rights Ireland, the NGOs Privacy International, la Quadrature du Net, and Big Brother Watch, who started these cases, acted not only as public watchdogs denouncing situations that conflicted with respect for the right to privacy and data protection, they also initiated strategic litigations that prompted major reforms at both EU and national levels. From a constitutional point of view, one can argue that civil society actors contribute to balancing powers, triggering a judicial reaction to legislators and governmental agencies’ abuses or inaction.
Second, academics, activists, and NGOs have been among the most active actors in developing and advocating new constitutional principles that specifically target surveillance practices. Empirical studies have counted more than 200 declarations issued by civil society actors that aim to articulate and define principles and values for the digital society. Footnote 13 These texts are generally known as Internet bills of rights and represent an interesting source to provide an analysis from a legal and, more specifically, a constitutional point of view. Indeed, Internet bills of rights are non-legally binding declarations made by a plurality of actors outside the institutionalized constitutional processes. If such an informal character makes them less interesting for black-letter law scholars, this feature undocks them from the limitations of political agenda and institutional remits, giving them an unprecedented possibility to propose innovative normative solutions to the challenges posed by mass surveillance practices. Footnote 14 From a sociolegal perspective, these non-legally binding declarations represent an invaluable source for understanding civil society’s claims in the context of digital society and the extent to which these demands are accepted and integrated into the legal order.
III. Internet Bills of Rights: From the Right to Privacy to the Prohibition of Mass Surveillance
Internet bills of rights’ task of articulating and defining principles and values for the digital society are not confined to a superficial conceptual make-up. These declarations do not merely specify the applicability of existing constitutional guarantees in the digital society. If one talks of a right to freedom of expression online, one instinctively grasps that communication through digital technology should be guaranteed. The same can be said for freedom and secrecy of correspondence, freedom of association, and peaceful assembly. The development of digital technologies provides new instruments to exercise these rights. The communicative power of these constitutional provisions is generally preserved because one instinctively considers new technologies as the heir to an old means of communication. In these cases, Internet bills of rights tend to simply restate the applicability of analog constitutional provisions in contemporary society by referring to digital communication tools.
However, a similar intervention of superficial stuccoing does not always suffice. A series of fundamental rights designed for an analog world struggles to permeate and inform virtual reality. Their current formulation does not immediately speak to the actors of the digital society. Individuals and public and private institutions do not instinctively grasp the implications of these fundamental values. In these cases, Internet bills of rights intervene to extract their meaning and significance for a contemporary digital society. These declarations distill the quintessence of these constitutional values. They unfasten the pure essence of these rights from the accidentals of the society in which they emerged. At this point, the ultimate telos of the constitutional norm is plunged into the context of the digital society, from which it resurfaces with a more specifically tailored layout. Teubner, dealing with the emergence of constitutional patterns beyond the state, similarly discusses a double process of “generalization” and “re-specification.” Footnote 15 The constitutional norm cannot be merely transplanted in a different societal context. One needs to understand its ultimate aim, generalize its principles, purify it from its original contextual contaminants, and then re-specify it in light of the characteristics of the new social reality.
The respect for individual privacy is an example of a fundamental value that Internet bills of rights do not merely restate with reference to digital technology but is generalized and further re-specified in light of the challenges of the digital society. Footnote 16 Most of these declarations of rights generally affirm that any form of surveillance affects the individual’s privacy. Footnote 17 However, some of them introduce new principles limiting mass surveillance practices. For example, in the words of the African Declaration on Internet Rights and Freedoms:
Mass or indiscriminate surveillance of individuals or the monitoring of their communications constitutes a disproportionate interference, and thus a violation, of the right to privacy, freedom of expression, and other human rights. Mass surveillance shall be prohibited by law. […] In order to meet the requirements of international human rights law, targeted surveillance of online communications must be governed by clear and transparent laws which, at a minimum, comply with the following basic principles: first, communications surveillance must be both targeted and based on reasonable suspicion of commission or involvement in the commission of a serious crime; second, communications surveillance must be judicially authorized, and individuals placed under surveillance must be notified that their communications have been monitored as soon as practicable after the conclusion of the surveillance operation; third, the application of surveillance laws must be subject to strong parliamentary oversight to prevent abuse and ensure the accountability of intelligence services and law enforcement agencies. Footnote 18
The constitutional message advocated by these civil society groups is clear. Internet bills of rights affirm that mass and indiscriminate surveillance represents an inadmissible limitation on the right to privacy, which cannot be considered proportionate as a matter of principle. Footnote 19 Only targeted forms of surveillance, subject to specific guarantees, can be tolerated as reasonable restrictions of that right. Footnote 20 Interestingly, guarantees such as the requirement of serious crimes, judicial authorization, and notification to individuals align with those progressively established by the CJUE case law, as we will see in the following section.
C. Judicial Activism: A Gradual Process of Constitutionalization
I. Indiscriminate Data Retention in the EU: Legislative Genesis and First Judicial Ban
As anticipated, the “great temptation” to exploit the vast potential of technological tools and sophisticated surveillance instruments significantly expanded in the EU in the aftermath of 9/11. Considering the importance of effectively facing terroristic threats, the generalized retention of communications data by telecommunication operators immediately appeared an invaluable measure, allowing intelligence agencies and law enforcement authorities to “go back in time” Footnote 21 and find useful investigative leads about people previously unknown to public authorities. Footnote 22 For these reasons, the so-called e-Privacy Directive 2002/58 Footnote 23 was adopted just a few months after the attack on the Twin Towers, followed by the Data Retention Directive (DRD), Footnote 24 which was rapidly implemented after the 2005 terrorist attacks in the EU. The case law, developed at the national and supranational level in the EU, concerning the implementation of data retention instruments and disciplines, represents a relevant, complex, and clear exemplification of the judicial inputs to the vast debate on surveillance tools and their constitutional limits. Analyzing the main arguments employed and the guarantees and safeguards affirmed by national and EU judges helps us better understand the consequent reactions and the normative solutions developed or still being discussed today by legislators.
Starting from the first adopted provision concerning data retention, Article 15 of the e-Privacy Directive establishes a broad derogation to the general obligation to erase or anonymize metadata. In particular, Member States can implement legislative measures allowing communications data to be exceptionally retained for a specific and limited time period when it is considered a “necessary, appropriate and proportionate measure within a democratic society to safeguard national security (i.e. State security), defense, public security and the prevention, investigation, detection, and prosecution of criminal offenses or unauthorized use of the electronic communications system.”
When the exacerbated terrorist threat pushed the EU legislators to increase and intensify harmonized answers in the context of a war on terror, the DRD, complementing the vague discipline provided by Article 15 of the e-Privacy Directive, Footnote 25 obliged Member States to implement national legislation requiring all service providers to retain a vast list of communications data for a period between six months and two years, attributing to the discretionary evaluation of each national legislator the adoption of specific access conditions and procedures as well as the definition of which “serious crimes” would legitimize the processing of metadata by law enforcement authorities. Footnote 26
Setting aside the challenges concerning the DRD’s legal basis (decided by the CJEU in Ireland v Parliament and Council), Footnote 27 the mass surveillance measures and obligations established by the Directive have been controversial from the outset. The complex legislative debate at the EU level and the doubts expressed by the Article 29 Working PartyFootnote 28 and numerous NGOs Footnote 29 had a significant impact. In several Member States, the domestic laws implementing the DRD were challenged before the national courts, which were asked to assess the compatibility of a bulk retention regime with national fundamental rights and principles. Footnote 30 In most cases, the relevant judicial decisions on mass surveillance and blanket data retention measures were declared to be a disproportionate intrusion into the private sphere. Notwithstanding the different approaches, Footnote 31 this case law revealed an important opportunity for national judges to deal with the profound issues of a generalized surveillance system, particularly the need for precise guarantees and limits to restrict intrusive measures.
Even though none of these national decisions considered the validity of the Directive itself, this piece of EU legislation was challenged before the CJEU soon after. What is now defined as the “data retention saga,” which is continuing nowadays, represents, together with the important judicial impulses coming from national courts, the most significant stimulus to the constitutional debate over data retention systems in the EU, promoting a complex process of constitutionalization of mass surveillance regimes in general. In fact, in the landmark Digital Rights Ireland decision, Footnote 32 the CJEU affirmed some crucial and unprecedented principles on mass surveillance systems. First, generalized and indiscriminate metadata retention, independently from potential subsequent access to the retained data by state authorities, was considered an invasion of users’ private life per se. The Court recognized that communications data, regardless of the content of the communication itself, may allow precise conclusions on users’ habits, everyday life, place of residence, daily movements, activities, relationships, and social environments, which can generate “in the minds of the persons concerned the feeling that their private lives are the subject of constant surveillance” (paragraph 37). By demonstrating a profound awareness of the possible risks and consequences related to the existence of massive retention systems, even if motivated by security purposes, the Court noted that mandatory bulk and generalized retention, covering all electronic communications services as well as the entirety of users and communications data produced, could not represent a measure limited to what is strictly necessary. The Court underlined that a form of legitimate and proportionate retention could only be identified in a targeted regime, which means a retention limitedly affecting people for whom there is evidence capable of “suggesting that their conduct might have a link, even an indirect or remote one, with serious crime” (paragraph 58), or specifically limited to a geographic area and/or a time period and at the unique purpose of preventing or fighting against serious crimes. The lack of these limitations, regarded as essential to ensure a legitimate and proportionate compression of individuals’ fundamental rights, determined the invalidation of the DRD.
By defining a clearer standard for the future development of mass surveillance techniques in the EU, the CJEU determined that security purposes, even when motivated by terrorist threats, cannot be placed outside the constitutional human rights guarantee framework established by national constitutional charters and supranational fundamental rights declarations. Footnote 33 In their decision, the EU judges affirmed some limits and conditions on surveillance mechanisms by prompting the adoption of targeted measures in clear contrast with the generalized dimension of massive surveillance systems.
With the invalidation of the DRD, Article 15 of the e-Privacy Directive once again became the only piece of EU law that regulated the possibility of retaining communications data for security purposes, with Member States soon adopting different approaches to data retention discipline. National courts were once again asked to evaluate the compliance of internal data retention regimes with national and EU fundamental rights, as interpreted by the CJEU. Footnote 34 Doubts and criticalities emerged again concerning competence matters and the proportionality of national bulk data retention systems. The attention and activism on this topic reflected the difficulty, also for national judges, in determining a correct balance between security needs, supposedly better guaranteed by bulk surveillance instruments, and protecting rights and freedoms.
In light of the different and fragmented reactions of national courts and legislators, several preliminary rulings were promoted by Swedish, British, and Spanish courts, all asking for clarifications as to the interpretation of Article 15 of the e-Privacy Directive. In the important Tele2 Footnote 35 and Ministerio Fiscal Footnote 36 cases, the CJEU reaffirmed the principles already established in Digital Rights Ireland, stressing that national provisions on data retention and access to retained communications data for security purposes fall entirely within the scope of application of EU law. In all these cases, in which the EU Court was first confronted with surveillance tools, the judges acted as constitutional “lawmakers,” giving very precise indications to national legislators, Footnote 37 prompting the incorporation of core constitutional values such as the rule of law, due process guarantees, and the presumption of innocence into the discipline regulating data retention. By explicitly declaring that Article 15 of the e-Privacy Directive, read in light of the CFREU, must be interpreted as precluding national legislation from providing general and indiscriminate retention of all traffic and location data of all users, including all means of electronic communication, the Court reiterated the importance of clear and precise rules for establishing a form of targeted data retention regime, together with strict rules and prior control regulating access to data by law enforcement or intelligence authorities. Footnote 38
II. Constitutional Fine-Tuning: Security and the Ban of ‘Systematic’ Mass Surveillance
Notwithstanding the important constitutionalization path of the data retention surveillance instrument promoted by the abovementioned CJEU case law, Member States demonstrated a reticent approach; they were reluctant to abandon bulk regimes perceived as invaluable and irreplaceable tools in the fight against national and transnational security threats. Footnote 39 The solution of targeted data retention, highly criticized for its limited efficacy and potential discriminatory impact, Footnote 40 was not implemented at the national level. Footnote 41 The vast majority of governments, legislators, and, in some cases, courts Footnote 42 adopted a “defensive” approach, trying to promote an interpretation of the CJEU case law that would not definitively ban generalized and indiscriminate retention systems but provide these regimes with stronger and more precise rules governing the access phase by state authorities. This erosion of the stringent principles declared in the data retention saga raised strong criticisms in civil society and NGOs. They considered national legislation that allowed mass surveillance instruments to be per se incompatible with the fundamental rights guaranteed by the CFREU as interpreted by the CJEU judgments on the topic.
In an impressive number of cases, Footnote 43 the divergent views expressed in front of national courts led to a request for the intervention of the CJEU, underlining how the limits and safeguards described in the previous case law produced controversial effects and required clearer explanations. In particular, governments, law enforcement authorities, and, in some cases, national courts asked the CJEU to “clarify, refine or even reconsider various aspects of its case law,” Footnote 44 arguing that strict safeguards and limits on access to data would be sufficient to balance the interference produced by generalized and indiscriminate retention. In all these decisions, EU judges pronounced unprecedented and crucial principles destined to influence the employment of mass surveillance techniques.
Having reiterated that mass surveillance techniques cannot be exempted from the control of EU law and the CJEU,Footnote 45 the judges strongly reaffirmed that bulk data retention for public security purposes – namely, the fight against serious crimes – cannot be considered proportionate and legitimate. By identifying targeted data retention as the only form of surveillance allowed, the Court refused to accept alternative retention regimes such as the restricted one Footnote 46 and decided not to succumb to the strong and profound criticism governments and law enforcement authorities expressed on the effectiveness and non-discriminatory nature of the targeted solution. If these positions represented a clear reaffirmation of the principles established in the previous case law, the most innovative part of these recent decisions – in particular those of October 2020 – was the one recognizing a sort of mitigation of the previous approach, highlighting the ongoing and still evolving nature of the process of constitutionalization performed by the Court. Protecting national security from serious threats such as terrorism or other activities capable of destabilizing a country’s fundamental constitutional, political, economic, or social structures provides Member States with wider maneuvering room. The seriousness of this objective may justify even more severe interferences with fundamental rights, including bulk data retention techniques. Footnote 47
Although this affirmation can appear as a major victory for governments and public authorities vocally invoking the possibility of adopting forms of massive surveillance and retention, the Court carefully and precisely specified all the limits and conditions to harness the use of such a highly invasive instrument. We can identify a further constitutionalizing effort by the CJEU, which introduced constitutional safeguards and boundaries by requiring national legislators to determine: (i) a precise legislative framework disciplining the possibility of adopting bulk data retention regimes; (ii) a specific time limit to the use of this instrument; (iii) rigorous data security safeguards able to protect from the risks of abuses; and (iv) control by courts or independent administrative bodies. Moreover, the Court specified that a bulk data retention regime could never be admitted in the context of the fight against serious crime, while it can be justified to face national security threats; However, even in this case, it cannot be systematic in nature, and its adoption has to be subordinate to the existence of serious, real, present, or at least foreseeable, threats to national security. Footnote 48 Although the guarantee of national security is considered vital, reaching this critical objective cannot be untied from respect for the rule of law, which notably is
characterized first and foremost by the requirement that power and strength are subject to the limits of the law and, in particular, to a legal order that finds in the defense of fundamental rights the reason and purpose of its existence. (…) If it simply gave primacy to mere effectiveness, a State based on the rule of law would lose that distinguishing quality and might, in extreme cases, itself become a threat to the citizen. If the public authorities were armed with a panoply of instruments of criminal prosecution such as to enable them to disregard or violate fundamental rights, there would be no way of ensuring that their uncontrolled and entirely unfettered actions would not operate ultimately to the detriment of everyone’s freedom. Footnote 49
The Court tried to establish a correct balance between the need to safeguard national and public security effectively and to ensure the respect of the democratic bases of our societies. This balancing exercise passed through a progressive constitutionalization of data retention regimes, through which the Court pushed national legislators and EU institutions towards adopting constitutional principles able to determine the limits of acceptability and the justifiability of fundamental rights’ compression for both public and national security purposes. Footnote 50
To conclude this overview of the CJEU’s contribution to the constitutionalization of mass surveillance practices in the EU, it is worth mentioning that the data retention saga did not represent the only case in which the CJEU was asked to deal with mass surveillance. The CJEU jurisprudence concerning the transfer of personal data collected in European territory and directed outside EU borders confirmed the profound impact on fundamental rights produced by massive collection, retention, and access to personal data operated by public authorities, even when motivated by security purposes concerning third countries. Footnote 51 In the Schrems I and Schrems II cases, the CJEU indirectly analyzed the intrusiveness of American mass surveillance systems such as Prism and Upstream by twice invalidating the adequacy decisions regarding data transfers to the US. Considering these provisions, the Court stated that “legislation permitting the public authorities to have access on a generalized basis to the content of electronic communications must be regarded as compromising the essence of the fundamental right to respect for private life, as guaranteed by Article 7 of the Charter,” Footnote 52 while a law not providing for any possibility for an individual to pursue legal remedies in order to have access to personal data relating to him, or to obtain the rectification or erasure of such data, does not respect the essence of the fundamental right to effective judicial protection (Article 47 CFREU).
In the Schrems saga, the Court tried to transpose the constitutional principles and guarantees already affirmed inside EU territory in cases wherein the surveillance is perpetrated by foreign authorities, with a complex and – from some aspects – questionable approach. Footnote 53 Similarly, in its Opinion 1/15, the CJEU invalidated the draft PNR agreement with Canada concerning transferring and processing EU passengers’ PNR data to Canadian law enforcement and security authorities. Here, the Court clearly reaffirmed the necessity of a connection or at least an indirect link between a possible threat and the retention, accessing, and processing of personal data, as well as the need for specific and precise safeguards guaranteeing the adequate protection of data outside EU borders. Footnote 54 The attention and relevance attributed to retention, concerning not only communications data but also other information such as PNR, clearly emerged from the numerous proposed preliminary rulings regarding the validity of the PNR Directive, Footnote 55 challenging its legitimacy based on the “constitutional” principles emerging from the data retention and Schrems saga. Footnote 56
D. Advancing in the Path of Constitutionalization? Legislative Inertia and Future Scenarios
The reiterated intervention of the CJEU in the field of data retention and data transfers demonstrates the profound questions and the complex debate caused by the use of mass surveillance techniques. Member States have reacted in very different ways to the ongoing path of constitutionalization launched by the EU judges, creating a fragmented landscape of solutions and approaches. For example, following the October 2020 decisions, the Belgian Constitutional Court Footnote 57 annulled national legislation regulating data retention, considering the bulk nature of that regime as fundamentally incompatible with the principles affirmed by the CJEU. The Belgian Government then proposed a reform to create a targeted data retention system based on geographical criteria, differentiated according to the threat authorities are asked to face; that is, different levels of danger associated with varying degrees of legitimate intrusion into the private sphere. Footnote 58 With this attempt, the Belgian Government demonstrated its willingness to renounce the bulk data retention instrument for public security purposes, trying to conform its national legislation to the constitutional principles and proportionality evaluations promoted by the CJEU data retention case law as well as confirmed by national courts – the Belgian Constitutional Court had already annulled a national data retention law in the aftermath of the Digital Rights Ireland decision. Footnote 59
Compared with Belgium, France adopted a very different approach. On April 21, 2021, Footnote 60 the Conseil d’État considered France’s national law providing a generalized data retention regime for national security purposes compliant with EU legislation and the CFREU – except for some specific provisions or omissions, such as the lack of a preliminary control by an independent authority and the lack of a periodic review aiming at verifying the persistence of serious, present, or foreseeable risks to national security. Embracing a permissive interpretation of the stringent requirements imposed by the CJEU, Footnote 61 the French judges recognized the impossibility of adopting a targeted data retention regime, as suggested by the CJEU, considering it unfeasible to determine in advance the possible geographical areas, subjects, or categories of data that could be useful for preventing or investigating serious crimes. These considerations appear in clear contrast with the solutions proposed by the Belgian government and the Belgian Constitutional Court, which clearly made efforts to go in the direction of the requirements indicated by the CJEU case law.
While the abovementioned landmark decisions adopted by the CJEU caused diverse legislative and judicial reactions all over Europe, there are countries where this case law has not produced a significant impact. In Italy, for instance, the debate on the proportionality and legitimacy of data retention was not fully considered by either the Italian legislator or the national courts. Only in recent times, after the CJEU decision in H.K. v. Prokuratuur, the Tribunale di Rieti decided to promote the first preliminary ruling related to data retention and access regime in Italy, Footnote 62 prompting an unprecedented but highly awaited multilevel dialogue between national and supranational courts. Footnote 63 This decision represents a significant novelty compared to the Italian Supreme Court approach, which had always considered the national data retention legislation perfectly compliant with the complex and highly discussed CJEU requirements. Footnote 64 Considering the intense debate generated by the approach promoted by the Tribunale di Rieti, the Italian legislature decided to intervene. In November 2021, a significant reform was approved by the parliament, Footnote 65 which modified the data access regime by introducing, for the first time, a prior control by a judge as well as a specific list of serious crimes legitimizing access to metadata. However, despite these changes, which promoted more stringent guarantees and ultimately greater compliance with the CJEU requirements regarding the access procedure, the Italian legislator decided not to reform the core architecture of its data retention regime. In Italy, current legislation establishes a retention period of seventy-two months without distinguishing between national security purposes or public security objectives and confirms a bulk and generalized data retention system. These provisions contrast with the specific limits established by the CJEU case law yet are still in place despite strong criticism by the Italian Data Protection Authority, academics, and experts. Footnote 66
The reactions and lively debate on the path of constitutionalization developed by the CJEU on data retention and mass surveillance, more generally, are still expanding. In April 2022, the Portuguese Constitutional Court declared many articles of the so-called metadata law,Footnote 67 adopted in accordance with the DRD provisions, unconstitutional.Footnote 68 It took several years for Portuguese constitutional judges to adopt such a decision, preceded by a lively political and judicial discussion that had brought, in the past, the same Court to express a very different position on the national data retention legislation. In Case no. 420/2017, the Portuguese judges (mainly) confirmed the legitimacy of bulk retention of basic metadata, for example, name and IP address. They relied on the Public Prosecutor’s Cybercrime Office affirmations, according to which the “Digital Rights ruling should not impair the bulk retention of metadata under Law 32/2008,” because “retention must be indiscriminate,” and “must include all citizens.”Footnote 69 Overcoming these positions, the 2022 decision opened a different and more careful approach to mass surveillance instruments in Portugal, prompting a serious debate on a renewed data retention legal framework and its proportionality. Once again, this example reveals, on the one hand, the key role played by the CJEU case law and, on the other hand, the difficulty for national legislators and Courts to promote a profound debate and normative process on how to constitutionalize surveillance tools that could incorporate fundamental rights protection and paramount constitutional principles successfully.Footnote 70
The varied but cautious and reluctant approaches adopted by Member States following the CJEU rulings in this field clearly emerge when analyzing the e-Privacy Regulation proposal, which aims to modify the existing and quite outdated e-Privacy Directive. The text, approved by the EU Council in February 2021 and currently subject to the legislative procedure, includes a specific provision Footnote 71 aimed at excluding from the scope of the application of the e-Privacy Regulation and EU law, more generally, all data processing procedures intended to protect national security, thus bypassing the application of the CJEUs stringent requirements, and potentially representing a stop to the path of constitutionalization described above. Footnote 72 The debate on adopting this new Regulation is still ongoing. It has encountered several difficulties and forced stops due to the lack of consensus between Member States and between different EU Institutions on some thorny issues. Data retention discipline is one of them.
The CJEU’s impulses on the surveillance debate in the EU and its Member States are questioned for different and opposing reasons. Several governments and legislators, sometimes even courts, considered the EU judges proportionality test on data retention too strict and the proposed limits and solutions, such as the targeted data retention, unfeasible or useless. On the contrary, as we have seen in previous sections of this work, human rights activists and experts pushing for a total ban on mass surveillance instruments looked at the more recent CJEU case law as a form of legalizing unlimited surveillance methods for national security purposes. Consequently, problems and uncertainties still persist in this field. As Podkowik and others put it, “both the EU Commission and the national legislatures have been struggling with the proper introduction of these standards into the EU and national legislation.”Footnote 73 Notwithstanding this relevant consideration, it cannot be denied that the efforts and inputs provided by the CJEU decisions, reinforced by the multilevel dialogue with national Courts, have determined the affirmation of a consistent and comprehensive path of constitutionalization intended to define how mass surveillance tools can be legally incorporated into a constitutional framework. Whether the more recent rulings represent a “progressive legitimation” of mass surveillance in certain cases or a “shift towards a more pragmatic approach,”Footnote 74 the CJEU intervention in many cases, even if with different degrees of efficacy, helped raising the level of fundamental rights protection by national legislators with regard to mass surveillance instruments.
Whether the current guarantees comply with the principles established by the EU judges is a question still open to discussion.Footnote 75 Nonetheless, in recent years, several national normative solutions proposed a serious attempt to constitutionalize mass surveillance through a path that Tzanou defined as “proceduralization.” This ongoing process and debate do not aim to prohibit mass surveillance per se but include constitutional guarantees into its normative discipline. The constitutionalizing process boosted by the rich case law on data retention should not be seen as limited to this particular surveillance tool. On the contrary, the importance of the principles affirmed should be likewise usefully applied to other instruments of massive surveillance, such as those based on artificial intelligence. Considering the firm push provided by this judicial intervention, legislators are now in charge of a difficult and delicate task: providing individuals with the constitutional safeguards that have progressively emerged from civil society demands and the vast case law described through clear legislative choices.
E. Conclusion
By establishing an intrusive form of control over people’s communications, the data retention regime and its regulation represent a central case study for an in-depth analysis of civil society and judicial impulses characterizing the constitutionalizing efforts in the field of mass surveillance. The opposition expressed by national security and enforcement authorities reveals a strong need to correctly balance the efficacy of investigative instruments and the guarantee of high-level security on the one hand and the protection of constitutional principles on the other. This prevents intrusive measures from eroding the trust between citizens and public authorities, which characterizes democratic societies based on the rule of law.
The multiple cases decided by the CJEU and the different approaches exhibited by Member States show how the challenges and issues linked to the use of surveillance instruments remain open and at the center of a complex and delicate debate that raises serious concerns. While not determining a rigid ban on every possible form of surveillance, at least for national security purposes, EU judges have tried to set precise limits to bulk data retention mechanisms by promoting a strong proportionality test. The clear voice of the CJEU, in contrast to the stunning silence of the European legislator, is now destined to leave space for the reactions of national courts, parliaments and the EU Commission. While the former should consider the need to modify national legislation regulating bulk data retention regimes, the latter must decide whether to promote a new ad hoc legislative proposal on data retention that prompts a homogeneous discipline backed by Member States, able to include the results of the process of constitutionalization promoted by the CJEU. Notwithstanding the difficulty of foreseeing possible future evolutions in this articulated field, the future of mass surveillance in the EU seems to be more predictable, thanks to the constitutionalizing efforts of civil society, the CJEU, and national courts in this context.
As evidenced by the impulses analyzed in this Article, the necessity to establish limits and conditions for vast surveillance instruments represents a complex challenge that does not have a solid, conclusive answer in practical terms. On the one hand, the role played by civil society acted as a significant stimulus to the development of an articulated set of judicial decisions related to the delimitation of mass surveillance systems. This highlights a profound awareness of the intrusiveness and impact of these instruments not only on the rights to privacy and data protection but also, more broadly, on the very relationship between citizens and public powers and between freedoms and controls. On the other hand, the questions of the relevance and effectiveness of bulk interceptions, retention, and access operations for security purposes make it hard to establish restrictive limits to the use of these tools in light of the seriousness of the threats they aim to face.
Notwithstanding specific demands from civil society actors, incorporating an explicit constitutional provision affirming an absolute prohibition of mass surveillance practices does not appear to be a feasible or realistic solution. Analyzing the combination of civil society and judicial impulses shows a more complex process of gradual constitutionalization of mass surveillance. Core constitutional principles and their corollaries, such as the principle of proportionality and strict necessity, the need to guarantee prior authorization and control by independent authorities, and the presumption of innocence imposing targeted approaches and objective criteria linking the surveillance measure to a real threat, are being progressively instilled at the regulatory level regarding surveillance instruments. The constitutionalization of mass surveillance does not imply a constitutional codification of an outright prohibition: a ban would eventually fail to strike a balance between the protection of state security and the compression of fundamental rights. Constitutionalizing mass surveillance in the EU is taking place by introducing specific limits and procedural guarantees derived from translating existing constitutional values in the context of the digital society.
This phenomenon is part of a broader multilevel and multistakeholder process of constitutionalization, progressively pushing the constitutional ecosystem to react to the challenges of the digital revolution. Footnote 76 The strengthening of this tendency and its effectiveness in relation to the constitutionalization of mass surveillance will depend on the simultaneous and interconnected work of various players, from civil society actors, who act as a watchdog and as a trigger of strategic litigation, to supranational and national courts and national legislators. This Article has shown that, after the significant efforts of these first two groups of actors, the question of the future implementation of legitimate mass surveillance systems now lies in the hands of legislators. The effective constitutionalization of mass surveillance practices implies the incorporation of the constitutional values and principles advocated by civil society and translated by case law into precise surveillance laws and regulations at both the national and EU levels.
Acknowledgements
We want to thank those who participated in the data protection and surveillance panel of the Annual BILETA Conference 2021 for their feedback on an earlier version of this Article. We want to thank Ms Cerys Lee for her research assistance. Giulia Formici would like to express her gratitude to the University of Milan, in particular to the Department of Italian and Supranational Public Law, where she held the position of Research Fellow until November 2023. While the Article has been elaborated jointly by the co-authors, Sections A and B were written by Edoardo Celeste, and Sections C and D by Giulia Formici.
Competing interests
The authors declare none.
Funding
The authors declare no specific funding.
Open access
