The objective of the Article, therefore, is to elucidate the key divergences between data and consumer protection with reference to the overlap between the GDPR and the UCT Directive to highlight the fundamental difficulties associated with the alignment of protections. Building on this analysis, the role of competition law will be examined. In particular, the inherent asymmetries associated with the practical implementation of pre-formulated declarations of consent and take-it-or-leave-it choices arguably point to the relevance of competition law analysis as a mechanism to ensure the availability of choice—as opposed to the protection of the ability to choose safeguarded by the consumer and data protection frameworks.
B. Aligned Citizen-Consumer Protections—Effectuating Meaningful Choice
The introduction of the Lisbon Treaty was a watershed moment for the protection of the fundamental right to data protection. Through the adoption of the Lisbon Treaty, the Charter of Fundamental Rights of the European Union (the Charter) was given binding force, and, as a result, the right to data protection was recognized as a distinct right for the first time.Footnote 8 The GDPR, as a secondary framework adopted under Article 16 TFEU, which provides a basis for the EU to adopt legislation for the protection of the right to data protection, further specifies the operation and protection of the right to data protection in particular, and rights and freedoms more generally when personal data are processed.Footnote 9 Personal data are as “any information relating to an identified or identifiable natural person (“data subject”).”Footnote 10 When combined with the definition of pseudonymizationFootnote 11 and the clarification regarding the interaction between these two definitions,Footnote 12 it is apparent that the personal data definition encompasses any data capable of singling out an individual.Footnote 13 A controller is defined as the natural or legal person “which, alone or jointly with others, determines the purposes and means of the processing of personal data”Footnote 14 and, a processor, as any natural or legal person “which processes personal data on behalf of the controller.”Footnote 15
The data protection framework is designed to counteract power and information asymmetries between controllers, processors, and data subjects, and to strengthen the position of data subjects relative to controllers.Footnote 16 In doing so, the framework attributes rights to data subjects and obligations to controllers—and processers—and provides for a clear separation in responsibilities with controllers processing data subjects’ personal data—with or without contracting the services of a processor, who hold merely a passive function—and with each entity easily distinguishable within the framework.Footnote 17 But, the suitability of the data protection framework has been repeatedly questioned given the emergence of the so-called big data environment and the datafication of everything.Footnote 18
I. The GDPR, Fairness, and the Regulatory Response to Technological Development
Key data protection principles such, as inter alia, data minimization, purpose limitation, security and confidentiality, and accuracy are all arguably problematic in terms of practical application.Footnote 19 Such criticism has manifested itself clearly in terms of the positioning of informational self-determination and control as a key rationale for the protection provided by the framework. This reflects the questions surrounding the capacity of data subjects to act in their own best interests and make autonomous decisions given the inherent power asymmetries in the data subject-controller relationship. In response to these concerns, the data protection framework has recently been reformed. Although the GDPR largely upholds the traditional regulatory approach, which was evident in Directive 95/46/EC,Footnote 20 there have been some notable developments in response to the rapid technological change. More specifically, the GDPR has explicitly introduced the principles of accountability and transparency—the notion of data protection by design and default—and moved towards a risk-based approach. As illustrated in Figure 1 below, there is a clear focus on accountability and transparency in Article 5 GDPR. Indeed, although both the accountability and transparency principles had implicitly played a role in the Directive 95/46/EC, this is the first time that these principles have been expressly provided for in an EU data protection legislative text. Moreover, although risk has always been important in data protection, the GDPR is an example of a risk-based legislative framework in that it places risk at its operative center, thereby affecting the interpretation of all rights and obligations contained therein.Footnote 21
These additions accentuate the importance of the fairness principle.Footnote 22 More specifically, given that the GDPR is an example of decentered regulation—namely by placing controllers in charge of how they comply and thus balance fundamentals and interests when personal data are processed—there is increased reliance on risk and accountability and, as a corollary, the obligation for controllers to process personal data fairly.Footnote 23
Although fairness has long been positioned as a key tenet of data protection law, it appears to have gained increased significance in the GDPR. As argued more extensively elsewhere, the fairness principle can be divided into two key elements, namely: Procedural fairness and fair balancing. Each of these elements is then made up of specific components, as illustrated in Figure 2 below.Footnote 24
In simple terms, both elements will run concurrently in the context of any given processing operation. Essentially, the elements manifest themselves in the balancing of rights and interests and mandate that controllers must take the rights and interests of data subjects into account—or in other words, not “ride roughshod” over the wishes of the latter.Footnote 25 The fairness principle also requires controllers to ensure that data subjects are informed and capable of exercising their right to data protection—thereby seemingly burdening controllers with an obligation to be mindful of data subject’s interests and capacities. This role for the fairness principle is manifested in ex ante and ex post fair balancing and procedural fairness safeguards which are evident throughout the provisions of the GDPR. The ex ante application of the fairness principle here refers to the rights and obligations which apply prior to personal data processing—the application of the conditions for lawful processing in Article 6(1) GDPR, for example—whereas the ex post safeguards relate to the rights and obligations which apply during personal data processing—like the application of data subject rights.
In this manner, the GDPR, as a secondary framework designed to protect rights and freedoms and the right to data protection in particular,Footnote 26 aims to satisfy the requirements for the limitation of rights in Article 52(1) of the Charter, read in conjunction with Article 8(2) of Charter as outlined above. Article 52(1) of the Charter specifies that:
[A]ny limitation on the exercise of the rights and freedoms recognised by this Charter must be provided for by law and respect the essence of those rights and freedoms. Subject to the principle of proportionality, limitations may be made only if they are necessary and genuinely meet objectives of general interest recognised by the Union or the need to protect the rights and freedoms of others.Footnote 27
Therefore, the GDPR is a secondary framework based on fairness checks and balances which aims to balance competing rights and interests in line with this proportionality and necessity test in Article 52(1) of the Charter where personal data are processed. As mentioned above, in the pursuit of this objective, the fairness principle transverses the entire operation of the framework. To clarify, in an ex ante sense, the processing of personal data requires one of the conditions for lawful processing contained in Article 6(1) GDPR to be satisfied. Regarding the provision of ISS, three of these conditions are specifically relevant, namely: Consent,Footnote 28 contract,Footnote 29 and legitimate interests,Footnote 30 as represented below in Figure 3.Footnote 31
The purpose of the processing, the means used to achieve this purpose, and the interests at stake will determine which of these conditions may be applicable. Consequently, each of the three conditions—as with the other conditions in Article 6(1) GDPR—plays a distinct and delineated role.Footnote 32 Importantly, the Article 29 Working PartyFootnote 33 has observed that where large amounts of personal data are collected in a commercial setting, consent will often be the only appropriate condition.Footnote 34 The fairness principle is key in the operation of each of these conditions.
II. Consent, Unfair Terms, and Fair Personal Data Processing
Although consent has always been important to the operation of the data protection framework—being specifically mentioned in Article 8(2) of the Charter—reliance on this condition as a meaningful means of legitimizing personal data processing has been repeatedly questioned. In short, bounded data subject rationality and the cognitive biases exposed by behavioral economics research have undermined the value of a reliance on consent in relation to its correlation to the data subject true wishes and understandings.Footnote 35 More specifically, the multiplicity of requests for consent and the resulting apparent dilution of its importance, the stickiness of default settings, market effects including lock-in, the complex legalese evident in privacy policies, and information overload have seemingly undermined the value of data subject participation and rendered consent increasingly difficult to apply in practice.Footnote 36 In response to these issues, the EU legislator has strengthened consent in the GDPR, thereby recognizing an increased role for the controller in ensuring the legitimacy of this condition for lawful processing.
Consent is defined in Article 4(11) GDPR as “any freely given, specific, informed and unambiguous indication of the data subject’s wishes by which he or she, by a statement or by a clear affirmative action, signifies agreement to the processing of personal data relating to him or her.”Footnote 37 Articles 4(11) and 6(1)(a) GDPR, in Figure 3 above, are then further specified in the conditions for consent in Article 7 GDPR, as represented below in Figure 4.
Article 7 GDPR is a key GDPR innovation designed to empower data subjects. Although the operation of Article 7 GDPR will be explored in detail later, for our current purposes, it is important to highlight how this provision manifests the application of the fairness principle. In particular, Article 7 GDPR appears to establish a burden of care on controllers regarding their responsibility to ensure that data subjects have been informed and understand the provided information. The controller is required to be able to demonstrate consent,Footnote 38 keeping in mind that, in assessing the freely given definitional condition, rendering access to the service conditional on consent may invalidate the reliance on consent.Footnote 39
Moreover, it is the controller’s responsibility to take the interests and limitations of data subjects into account vis-à-vis the requirement for specific unambiguous information in line with the transparency principle. This is a direct manifestation of the procedural fairness element—in other words, controllers are burdened with care. In addition, it should be understood that, in the determination of the legitimacy of consent and thus the operation of the key definitional conditions,Footnote 40 there is also a key role for the fair balancing element, and, in particular, the operation of the freely given criterion. Put briefly, consent is supposed to represent a meaningful choice as evidenced by the ability to withdraw consent in Article 7(2) GDPR, but also the separation of contract and consent in Article 7(4) GDPR. In this vein, controllers are required to take the rights and interests of data subjects into consideration in order to ensure that the consent is legitimate.
Therefore, Article 7 GDPR, is a significant addition and has resulted in an intense debate surrounding the suitability of other conditions, namely, Article 6(1)(b) GDPR and Article 6(1)(f) GDPR, to legitimize commercial processing operations. As alluded to in the Introduction, the actions taken by Max Schrems against Google, Instagram, WhatsApp, and Facebook refer to this need for the separation of consent from the processing of personal data necessary for the provision of the service as provided for in Article 7(4) GDPR. Nevertheless, this separation is complicated by the cross-reference to the UCT Directive in Recital 42 GDPR regarding the fairness of pre-formulated declarations of data subject consent. Indeed, in addition to the entirely internalized fairness assessment—through the operation of the GDPR’s fairness principle—one must also take the UCT Directive into account. Recital 42 GDPR states that:
In accordance with Council Directive 93/13/EEC a declaration of consent pre-formulated by the controller should be provided in an intelligible and easily accessible form, using clear and plain language and it should not contain unfair terms. For consent to be informed, the data subject should be aware at least of the identity of the controller and the purposes of the processing for which the personal data are intended. Consent should not be regarded as freely given if the data subject has no genuine or free choice or is unable to refuse or withdraw consent without detriment.Footnote 41 [Emphasis added]
This recital, therefore, makes an explicit cross-reference to consumer protection law and more particularly stipulates the requirement that pre-formulated declarations of consent must respect the protections against unfair contractual terms in the UCT Directive. Similar to the GDPR, the UCT Directive works from the assumption that there is an imbalance in bargaining power between suppliers and consumers, and in essence provides that unfair terms shall not be binding for the consumer.Footnote 42 Unfairness under the UCT Directive consists of a substantive element—including good faith and significant imbalance components to be assessed at the national level—and a formal transparency and information provision, as provided for in Articles 3–5 of the Directive.Footnote 43 As noted by Donnelly and White, “[a]lthough the first of these mechanisms has inevitably been the more high profile, it is arguable that the primary weighting of Directive 93/13 is toward the latter mechanism.”Footnote 44 This observation reflects the strongly held national contract law traditions, which will be discussed in Section D, and also the traditional objections to regulatory inference with the notion of freedom of contract.Footnote 45
This cross-reference, however, has arguably muddied the waters between consent and contract due to the reliance on the UCT Directive in relation to pre-formulated declarations of consent when contract is provided for as a distinct condition for personal data processing in Article 6(1)(b) GDPR. In addition, this cross-reference raises questions regarding the precise overlaps between the respective notions of fairness contained in the frameworks, given that the UCT Directive does not have the same fundamental rights foundations as the GDPR. Indeed, in contrast to data protection, which, as noted above, is protected as a distinct fundamental right in Article 8 of the Charter, consumer protection is merely provided for as a principle in Article 38 of the Charter. In simple terms, rights and principles are weighted differently in terms of their significance.Footnote 46 As a consequence, plotting the relationship between the GDPR and the UCT Directive is not straightforward. Intuitively, the fact that the GDPR applies in a commercial context denotes that it plays an important role in the protection of consumer interests. But, the fundamental rights foundations of the GDPR brings such a simplistic conclusion into question. As a consequence, precisely mapping the relationship between the UCT Directive, as a specific B2 C framework, and the GDPR, as an omnibus regime designed to protect the fundamental rights of citizens—thereby extending beyond the mere B2 C context—in the protection of citizen-consumers is of key importance. With such apparent distinctions between the two instruments, the analysis now turns to a more substantive discussion of the overlaps.
C. Pre-Formulated Declarations of Consent and the Bits In Between
In the binding provisions of the GDPR, the potential for pre-formulated declarations of consent is not explicitly mentioned. Indeed, although Article 7(2) GDPR refers to the separation of data subject consent given in the context of a written declaration from the other matters which may be included in such a declaration, like in Figure 4 above, this provision remains neutral in terms of its origin and nature. Instead, Article 7(2) GDPR stipulates more generally that such a written declaration must be presented “in an intelligible and easily accessible form, using clear and plain language.”Footnote 47 Such an approach is also reflected in Article 12(1) GDPR, which states that:
[T]he controller shall take appropriate measures to provide any information referred to in Articles 13 and 14 and any communication under Articles 15 to 22 and 34 relating to processing to the data subject in a concise, transparent, intelligible and easily accessible form, using clear and plain language, in particular for any information addressed specifically to a child. The information shall be provided in writing, or by other means, including, where appropriate, by electronic means. When requested by the data subject, the information may be provided orally, provided that the identity of the data subject is proven by other means.Footnote 48
Hence, Article 12(1) GDPR also has a wider scope of application than merely pre-formulated declarations, as evidenced by the use of the words “or by other means.”Footnote 49 That being said, such mechanisms are certainly included within its scope. As clarified in the previous Section, although the GDPR’s binding provisions do not delineate between boilerplate and individually negotiated declarations, Recital 42 GDPR deals with the fairness of pre-formulated declarations with reference to the UCT Directive in the application of informed data subject consent as a condition for lawful processing.
I. Pre-Formulated Declarations of Consent
The specification in Recital 42 GDPR that pre-formulated declarations of consent are to be provided “in an intelligible and easily accessible form, using clear and plain language,”Footnote 50 repeats the terminology used in Article 7(2) GDPR and Article 12(1) GDPR, and also seemingly echoes the formal fairness element in the UCT Directive. Article 5 UCT Directive states that, “[i]n the case of contracts where all or certain terms offered to the consumer are in writing, these terms must always be drafted in plain, intelligible language. Where there is doubt about the meaning of a term, the interpretation most favourable to the consumer shall prevail….”Footnote 51 Given the clear overlap in terminology, it is pertinent to analyze the interpretation of these notions in the UCT Directive with reference to the Court of Justice case law.
Article 5 UCT Directive strongly relates to the principle of transparency with the Member States being explicitly required to include the transparency principle in their implementations with mere references to the Court’s established practices regarded as insufficient.Footnote 52 Micklitz has noted in his analysis of the UCT Directive that transparency can be categorized as a sub-category of good faith via the principle of legitimate expectations which stems from the plainness and intelligibility standards.Footnote 53 The meaning of plain, intelligible language was analyzed by the Court of Justice in the Kasler judgement where the Court placed this requirement within the broader setting of providing information before consumers are bound by a contract.Footnote 54 It further stipulated that this requirement is not limited to grammatical intelligibility but must instead be understood broadlyFootnote 55 in order to allow the consumer “to evaluate, on the basis of clear, intelligible criteria, the economic consequences for him which derive from it.”Footnote 56 Plainness appears to relate to a term’s legal effect including its consequences vis-à-vis ambiguous formulations and the requirement that such terms should not put the seller or supplier in an advantageous position.
In contrast, intelligibility incorporates a linguistic element in terms of legibility whereby, if the seller is aware or should have been aware—for example, if they had exercised reasonable care—that a term is linguistically unintelligible for a consumer, then the seller is required to ensure its intelligibility. This is particularly significant in relation to standard form contracts which must, according to this requirement, be designed plainly both optically and also in terms of editing.Footnote 57 In saying this, however, it should be noted that intelligibility inherently incorporates a qualitative aspect as well, in that the information provided must also accurately and adequately inform the consumer in order to facilitate informed consumer decision-making. More simply, making the terms legible should not result in a loss of nuance in relation to the nature of the contractual agreement regarding the rights and obligations contained therein. Consequently, the interpretation of the terminology in the UCT Directive that is common across the frameworks—plain and intelligible—appears to reflect the aims of the equivalent provisions in the GDPR. In saying this though, it is important to highlight the subtle differences in construction, but also the additional reference to the requirement that declarations of consent be drafted in an easily accessible form provided for in the GDPR. This requirement seems to further specify plainness and intelligibility as understood under the UCT Directive. In addition, one should also be aware of the insertion of the obligation for concise and transparent information and communications in Article 12 GDPR.
These differences raise three important observations. First, the addition of the terms concise and transparent in Article 12(1) GDPR appears to reflect both the inherent aims of the UCT Directive but also the important positioning of the transparency principle in the GDPR. This is also reflected in the further specification of the intelligibility requirement and thus the reference to the potential use of icons in the operation of the information requirements contained in Articles 13 and 14 GDPR.Footnote 58 Second, the use of the formal fairness element from the UCT Directive—both ex ante and ex post is indicative of the overarching role of the procedural fairness element in the GDPR. In particular, the use of this terminology and its application to both the information provision requirementsFootnote 59 and any communication to the data subject,Footnote 60 show that the operation of the procedural fairness requirement extends beyond merely pre-formulated declarations of consent. This contrasts with the purely ex ante requirements in the UCT Directive. Finally, the specification in Article 12(1) GDPR that the information requirements must consider if the data subject is a child reflects the contextual nature of the procedural fairness element. Therefore, pre-formulated declarations depend on their intended use, and this is indicative of the fair balancing element in data protection fairness and hence, the requirement for data protection by design and by default in Article 25 GDPR.
In addition to the above comparisons, it is also important to note that Article 5 UCT Directive also provides for the in dubio contra proferentem rule whereby if a doubt exists in terms of meaning of a contractual term, the most favorable interpretation for the consumer must prevail. As noted by Rott however, there is a degree of uncertainty regarding the practical operation of this principle in that:
[I]f in case of an intransparent term one chose the consumer-friendly interpretation to start with, the term may not be held to be unfair. If, in contrast, the term was first tested for its fairness in its consumer-unfriendly interpretation, it might be unfair and therefore invalid; which would most likely benefit the consumer more than a consumer-friendly version of it.Footnote 61
For our current purposes, it is important to note that it is unclear how this rule would play out in the data protection context given that personal data can be used for multiple purposes—for example, provided one of the conditions contained in Article 6(1) GDPR is satisfied, see Figure 3. More specifically, ambiguity in the terms relating to the use of personal data will largely be context-dependent but will also on the face of it be contrary to the requirements contained in the GDPR. The transparency principle is of key importance and controllers are required to provide accurate information. As noted by the Article 29 Working Party, controllers are required to provide full and specific information even if less information would be easier for the data subject to understand.Footnote 62
Consequently, ambiguity in data protection would be seen as a violation of the transparency, fairness, and accountability principles. Controllers are required to take data subject rights and interests into account in the application of fair balancing and are also attributed the burden of proof in demonstrating their compliance with this requirement in line with procedural fairness—and more generally the burden of care in the application of fair balancing. For instance, as noted above in Section B(II), one can refer to Article 7(1) GDPR and the requirement to be able to demonstrate that the data subject has indeed consented. This is also indicative of the third component in the procedural fairness element, namely timeliness, as manifested in the requirement for controllers to reply without undue delay to data subjects’ requests for information under Articles 15–22 GDPR. Therefore, the formal element of the UCT Directive must be viewed in tandem with fairness in the GDPR. In discussing Recital 42 GDPR, Svantesson observes that this provisions indicates that the GDPR should be understood as providing lex specialis guidance as to how unfairness in the UCT Directive should apply in the data protection setting.Footnote 63 Although there appears to be certainly merit to this observation at first glance, the substantive delineation in terms of protections as evidenced by recent enforcement actions may place such an observation in doubt. Instead, therefore, this reference may suggest that, rather than providing a confirmation of a lex specialis relationship, Recital 42 GDPR in fact indicates parallel, concurrent, but substantively distinct, fairness assessments, thus reflecting the differences vis-à-vis the Charter foundations of the respective frameworks as previously described. This contention will now be bolstered through an analysis of the substantive fairness element in the UCT Directive below.
II. Delineating Pre-Formulated and Individually Negotiated Terms
From the above, the formal fairness element in the UCT Directive appears to provide some interpretive guidance for the validity of pre-formulated declarations of consent. At the same time, it is to be placed alongside the operation of the data protection fairness principle and the GDPR as a secondary framework more generally.Footnote 64 That being said, however, aside from this reference to the formal fairness element of the UCT Directive, Recital 42 GDPR also inserts a specific reference to the requirement that such pre-formulated declarations should not contain unfair terms. Hence, the recital seemingly refers to the substantive fairness element in the UCT Directive.
1. Giving Meaning to Unfair Terms
The substantive fairness element in the UCT Directive contains both good faith and significant imbalance components. According to Article 3(1) UCT Directive, a term will be regarded as unfair “if, contrary to the requirement of good faith, it causes a significant imbalance in the parties’ rights and obligations arising under the contract, to the detriment of the consumer.” There has been intense academic debate surrounding the meaning of good faith and significant imbalance. More specifically, as contracts are rarely contrary to the interests of traders, the potential openness of the test has been criticized.Footnote 65 In addition, good faith was somewhat alien to the common law tradition.Footnote 66 Although at first the UCT Directive had little real impact, in more recent years there has been a dramatic increase in the number of references from the national courts, thus facilitating the emergence of a deeper understanding of the Directive’s substantive fairness element.Footnote 67
More specifically, in endorsing the opinion of Advocate General Kokott in the Mohammed Aziz case, the Court of Justice found that in the determination of whether there is a significant imbalance the national court should take the national rules that would apply in the absence of a contractual agreement into consideration.Footnote 68 The Court further observed such an analysis would allow the national court to assess the extent to which the contract results in a less favorable legal situation and that such an assessment should include an examination of the legal situation of the consumer with regard to the means at their disposal under national law to prevent the continued use of unfair terms.Footnote 69 In the determination of the circumstances in which such an imbalance arises contrary to the good faith requirement, the Court of Justice stated that Recital 16 UCT Directive should be considered. In particular, in addition to indicating that good faith may be satisfied if the seller or supplier deals fairly and equitably with the consumer by taking their interests into account, Recital 16 UCT Directive further provides that the assessment of good faith necessitates the consideration of the bargaining power of both parties—including the specific consumer vulnerabilitiesFootnote 70; whether the consumer was induced to accept the term; and also whether the goods or services were being sold or supplied by the special order of the consumer. As a consequence, the national court is required to assess whether the seller or supplier, dealing fairly and equitably with the consumer, could reasonably assume that the consumer would have agreed to such a term in individual contract negotiations taking the particular circumstances of the case into account.Footnote 71
The Court of Justice has subsequently confirmed the judgement in Mohammed Aziz and found that a significant imbalance does not necessarily have to relate to an economic disparity but instead can also be a consequence of a “sufficiently serious impairment of the legal situation in which that consumer, as a party to the contract, is placed, vis-à-vis a restriction of rights or a constraint on the exercise of such rights.”Footnote 72 Importantly, in this context it is also necessary to examine Article 4(1) UCT Directive which stipulates that:
[T]he unfairness of a contractual term shall be assessed, taking into account the nature of the goods or services for which the contract was concluded and by referring, at the time of conclusion of the contract, to all the circumstances attending the conclusion of the contract and to all the other terms of the contract or of another contract on which it is dependent.Footnote 73
This provision highlights the importance of the individual circumstances of each case and, therefore, the significant role played by the national courts as the evaluators of the fairness of a specific term, given that the Court of Justice does not normally have access to the full facts of the case.Footnote 74 In this context, the role of the annexed grey-list, which provides an indicative list of unfair clauses, should also be acknowledged.Footnote 75 Indeed, although the Court of Justice has stated on several occasions that the adoption of this list is up to the Member States and that there is no presumption of unfairness unless otherwise provided by national law, according to the Invitel judgement the inclusion of a term in the list remains an essential element on which the national court can base its assessment.Footnote 76
With this in mind, it is important to note that the UCT Directive is a product of its time in that in effect it amounts to a partial harmonization of consumer contract law, largely focused on information provision and transparency as adopted under Article 100a EEC (now Article 114TFEU). This Treaty provision forms the legal basis for the EU to adopt legislation for the approximation of laws for the functioning of the internal market. To clarify, Article 114 TFEU permits the EU legislator to regulate areas which are seen as obstacles to the proper functioning of the internal market. It should also be noted that, given the fact that the Directive is a minimum harmonization instrument,Footnote 77 Member States are not precluded from offering a higher level of protection. Establishing precise indicators for when a term is contrary to good faith is, therefore, a determination which remains challenging at the EU level. As a consequence, there is a large degree of disparity amongst the Member States with many opting to expand the protections—for example, if the existing protections were not already more expansive.Footnote 78 Accordingly, in order to truly assess the fairness of a particular term one is required to refer to the national courts.Footnote 79 The room for maneuver afforded by the minimum harmonization approach is illustrative of the tight reign that Member States have kept over national contract law, for more see Section D below. This is indicative of the Directive’s restricted scope in that it only concerns terms that have not been individually negotiated.
From the above, therefore, one must question how the substantive fairness element in the UCT Directive overlaps with the principle of fairness in data protection in order to interpret unfair terms in the context of pre-formulated declarations of consent. The need for such an assessment is particularly clear given that the UCT Directive focuses on economic considerations in comparison to the fundamental rights focus of the GDPR. In this vein, one can refer to the recent EDPS opinion on the proposed Directive on certain aspects concerning contracts for the supply of digital contentFootnote 80 in which the EDPS criticizes the use of Article 114 TFEU as a legislative basis concerning matters involving personal data due to the availability of Article 16 TFEU. Arguably this points to a larger issue with relying on frameworks based on Article 114 TFEU, and, thus, their suitability to cater for social rights orientated concerns rather than the market integration centered mandate understood more readily to be the focus point of Article 114 TFEU. To clarify, this is not to suggest that Article 114 TFEU cannot be used to harmonize protections but rather to question how this might affect the protection goal.
Although the relationship between Article 114 TFEU and Article 16 TFEU is a matter requiring more detailed analysis, for our current purposes it is sufficient to conclude that the precise effect of such considerations remains to be seen. But, given the potential distinction, it is necessary to question the overlap between the reference to the prohibition of unfair terms and the last sentence in Article 7(2) GDPR which, as noted above, states that, “[a]ny part of such a declaration which constitutes an infringement of this Regulation shall not be binding.”Footnote 81 Therefore, although it is clear that any part of a declaration, pre-formulated or not, which infringes the GDPR will not be binding, it is uncertain how this clarification differs substantively from the UCT Directive fairness assessment of pre-formulated declarations. As mentioned in the previous sub-Section, Svantesson suggests that the reference to the assessment of unfair terms under the UCT Directive could be viewed as providing lex specialis specification of the rules therein through the requirements in the GDPR.Footnote 82 Due to the fundamental deviations underlying the frameworks however, one should question such a conclusion. In support of this, one can refer to the common position taken by various national consumer protection agencies through the Consumer Protection Collaboration Network regarding the terms of service of social networking sites which clearly focused on more traditional cross-border consumer contract issues. These issues include clauses relating to jurisdiction, the identification of commercial communications, the waiving of liability, the removal of content and unilateral rights to change, determine the scope of and terminate agreements.Footnote 83 Accordingly, instead of incorporating an assessment of the fairness of terms in line with data protection—which would relate more to the application of the transparency principle and the validity of the data subjects’ consent—the common position focuses on issues more aligned with traditional B2 C cross-border contract issues. Such a distinction allows for the differentiation in the respective frameworks’ intent and their underlying policy objectives. This also appears to be reflected in the judgements of—at least some—national courts where national law contract formation requirements are satisfied.
In addition to arousing the interest of data protection authorities, however, the developments since the Facebook-WhatsApp merger have also caught the eye of consumer protection authorities. More specifically, the Federation of German Consumer Organizations sought an injunction to stop the continued data-sharing and the deletion of the data already transferred to Facebook.Footnote 92 Furthermore, on May 11th, 2017, the Italian Competition and Consumer Protection Authority (Autorita’ Garante della Concorrenza e del Mercato (AGCM)) adopted two decisions in its proceedings against WhatsApp.Footnote 93 Nevertheless, in line with the common position taken by a number of national consumer protection agencies, the AGCM rulings appear to diverge from the approach taken by the Article 29 Working Party in relation to the information requirements and, thus, the expectations of the respective enforcement agencies—thereby indicating deviations in substantive requirements. The action taken by the Article 29 Working Party focused on the failure to inform the data subjects of the sources and categories of data and to sufficiently specify the intended purposes. The Article 29 Working Party viewed these failures to be in violation of the transparency principle in the GDPR. In contrast, similar to the common position adopted by national consumer protection agencies, the AGCM, in its proceedings concerning the UCT Directive, focused instead on the fairness of clauses regarding issues such as inter alia the choice of jurisdiction and law clauses, the unilateral authority to interrupt the service without reason or advance notice, and to rescind or terminate the contract.Footnote 94
From the diverging interpretations by the relevant enforcement authorities, the formulation of Recital 42 GDPR, and the specific reference to the limitation in scope to pre-formulated declarations, it can be concluded that the Recital aims to clarify that pre-formulated privacy policies may be “unfair” under the distinct yet complementary UCT Directive protections. It is suggested, therefore, that the UCT Directive and the GDPR operate in an independent but complementary manner. This appears to be indicative of the bolstered protections for data subject consent, the common reliance on pre-formulated terms—especially in the ISS context—and hence, the apparent need to protect data subject-consumers from consenting to terms that they would otherwise not have agreed to had they been individually negotiated.
2. Pre-Formulated Versus Individually Negotiated and the Importance of Price
From the above, it is important to reiterate that the UCT Directive is a minimum harmonization instrument and is hence reliant on national law implementations, which can offer a higher degree of protection. This approach has consequently led to a large degree of disparity across the Member States, which is a reflection of diversity in terms of legal traditions and approaches in national contract law. Furthermore, this also reflects the focus of the UCT Directive on terms not individually negotiated. Indeed, as per Article 3(2) UCT Directive, terms drafted in advance are always considered as not being individually negotiated and where there is doubt the burden of proof rests with the seller or provider. As previously mentioned, this focus on pre-formulated terms reflects the underlying assumption that consumers who actually engage in negotiations with traders are protected from risk. Weatherill observes, however:
[T]his is by no means uncontroversial. One might go so far as to adopt precisely the opposite perspective and argue that face-to-face discussion deepens the risk that the economically powerful trader will exploit the consumer. However, the Directive’s limitation to terms that have not been individually negotiated demonstrates a suspicion of “mass-produced” contracts, at least at the threshold of jurisdiction to check enforceability.Footnote 95
The delineation of the UCT Directive’s scope is important, given that it reflects the ongoing importance of national law and courts, which are of substantial significance in the shaping the application of the protections.
Although it is hard to imagine an individually negotiated B2 C term especially in the ISS context, this does raise interesting questions in terms of practical application. Would the negotiation of any part of a contract render it pre-formulated or individually negotiated? And further, how are terms classified as pre-formulated in practice? In this regard, one can refer to Article 3(2) UCT Directive which stipulates that:
[A] term shall always be regarded as not individually negotiated where it has been drafted in advance and the consumer has therefore not been able to influence the substance of the term, particularly in the context of a pre-formulated standard contract.
The fact that certain aspects of a term or one specific term have been individually negotiated shall not exclude the application of this Article to the rest of a contract if an overall assessment of the contract indicates that it is nevertheless a pre-formulated standard contract.
Where any seller or supplier claims that a standard term has been individually negotiated, the burden of proof in this respect shall be incumbent on him.Footnote 96
In commenting on this provision and hence, the notion of individual terms, Micklitz argues that the term pre-formulated does not have to be interpreted strictly.Footnote 97 Micklitz substantiates this observation with reference to two important points. First, contracts between private persons based on individual negotiations are not included within the scope of the UCT Directive, and second, the core terms relating to the “price/quality ratio and the main subject matter of the contract” are not subject to review.Footnote 98
The first of these points inherently points towards the personal scope of the Directive and, thus, the specification in Article 1(1) UCT Directive that the Directive targets contracts concluded between a seller or supplier and a consumer. Regarding the second of these points, it is important to highlight Article 4(2) UCT Directive. This provision states that an:
[A]ssessment of the unfair nature of the terms shall relate neither to the definition of the main subject matter of the contract nor to the adequacy of the price and remuneration, on the one hand, as against the services or goods supplies [sic] in exchange, on the other, in so far as these terms are in plain intelligible language.Footnote 99
Consequently, the exemption of these core terms restricts the application of the substantive fairness element to the more peripheral contractual aspects—provided such terms are stipulated in plain intelligible language and in line with the formal fairness element—hence seemingly minimizing the impact on freedom of contract.Footnote 100 The key point which emerges from Article 4(2) UCT Directive, therefore, is that the Directive does not wish to control the fairness of the “price.”Footnote 101 In the context of pre-formulated declarations of consent, the question thus becomes: What constitutes a price? This is a matter for the national courts to determine with reference to the specific facts of the case. It is a hotly contested topic, given that distinguishing the specific limits of this notion remains challenging. More specifically, the case law on this issue has highlighted the stickiness of this concern,Footnote 102 and this has resulted in some disparity in interpretation between the national courts and the Court of Justice.Footnote 103
It should be noted, however, that even if a term under assessment is deemed to fall within the exemption provided in Article 4(2) UCT Directive, it is still subject to an overarching transparency requirement given that this same provision mandates that such terms be presented in plain intelligible language.Footnote 104 Given the focus of the present Article, however, it is necessary to explore this issue further. Hence, the question becomes one of how the exclusion of the “adequacy of the price and remuneration, on the one hand, as against the services or goods supplies [sic] in exchange, on the other,”Footnote 105 may affect the application of the UCT Directive in the context of pre-formulated declarations of consent. This raises an important fundamental sub-question: Are personal data to be considered a “price”? Such a finding would exempt the provision of personal data from the substantive fairness test, which would appear odd given the aim of pre-formulated declarations of consent. Although the ability to offer higher levels of protection remains and, thus, extends to the possibility of excluding the exemption provided in Article 4(2) UCT Directive in the national implementation,Footnote 106 the lack of a harmonized approach runs contrary to aims of the GDPR if personal data are positioned as the price.
Building on this discussion, one can refer again to the above delineation in the substantive application of the assessment of unfair terms in the UCT Directive and the fairness principle in the GDPR. Given that certain consumer protection authorities appear to position personal data as a de facto price and thus a core term, the substantive fairness element in the UCT Directive—at least unless otherwise provided for in the national implementation—does not apply in the enforcement of the Directive’s protections.Footnote 107 Importantly, this does not affect the application of the GDPR’s fairness principle. Although it will be discussed in more detail below, it is worth noting that this nuanced point perhaps adds further clarification in terms of delineating the substantive application of the fairness protections in the UCT Directive and the GDPR. Nevertheless, it is important to re-emphasize that this will depend on the national implementations and thus on whether national transpositions of the UCT Directive, or indeed the prior existing law, extend the substantive protections to core terms. As noted by Helberger, Borgesius, and Reyna, therefore, consumer law could be positioned as an important instrument in the assessment of the fairness of the conditions under which consumers agree to personal data processing.Footnote 108 Nevertheless, due to the minimum harmonization approach in the UCT Directive and the more specifically tailored rules in the GDPR, there is potential for disparity. It is thus argued that a violation of consumer law would merely result in the addition of supplementary enforcement mechanisms, rather than a further tailoring of the data protection requirements, as any breach of the GDPR will not be binding in line with Article 7(2) GDPR. Nonetheless, this is certainly not a straightforward matter.
As described above in Section B(I), the right to data protection aims to protect individual control over personal data. In its construction, this fundamental right recognizes the benefits of personal data processing but also aims to mitigate this by targeting the prevention of disproportionate impacts on individuals.Footnote 109 This is reflected in the triangular structure of Article 8 of the Charter—for example, controller obligations, data subject rights and the monitoring activities of the authorities.Footnote 110 Data subject control, however, remains key in the operation of the GDPR, given that the framework aims to not only protect fundamental rights and freedoms in general, but also the right to data protection in particular where personal data are processed in Article 1(2) GDPR. That being said, personal data are seen by many as the currency through which ISS are provided with certain consumer protection authorities clearly positioning personal data as a price.Footnote 111 Furthermore, the UCT Directive inherently assumes that some form of price will be exchanged in the operation of its provisions as evidenced by the exclusion of the core terms from the substantive fairness element assessment at the EU level. Therefore, the determination of what will be classified as a price is of clear importance. The following Section aims to more specifically tackle the categorization of personal data as a price, given that it is a particularly thorny issue in the alignment of the protections offered by the UCT Directive and the GDPR, and as it has been at the root of the reforms of the consumer law framework.
D. The Economic Value of Personal Data and Fair Personal Data Processing
The need for some form of value exchange or price is indicative of the fact that in order to assess the validity of the contract formation, one is required to refer to the national level. This reflects the failed attempts to harmonize contract formation at the EU level. More specifically, the jettisoning of large parts of the Consumer Rights DirectiveFootnote 112 during the negotiationsFootnote 113 and the failed Regulation of the European Parliament and of the Council on a Common European Sales Law—known as the Optional Instrument—are evidence of how controversial the harmonization of contract formation has been in practice. This is of particular relevance for the current analysis as the European Commission proposed the Digital Content DirectiveFootnote 114 as a means of filling the gap left by the failure of the Optional Instrument via a dilution of the Regulation’s ambitions—thus leaving the laws governing contract formation in the hands of Member States—instead aiming to recognize that data, including personal data, can be positioned as a form of payment. In support of this contention one can refer, for instance, to Article 5(b) of the failed Optional Instrument which aimed to recognize the validity of “contracts for the supply of digital content whether or not supplied on a tangible medium which can be stored, processed or accessed, and re-used by the user, irrespective of whether the digital content is supplied in exchange for the payment of a price.”Footnote 115
In simple terms, this failed proposal aimed to recognized that there was no need for a price in order for a contract to be formed. In essence, by proposing the Digital Content Directive the Commission wished to avoid the problems of the past associated with harmonizing contract formation at the EU level, and instead aimed to extend protections to consumers in situations where personal data are effectively used as the means of payment. Such changes are also manifested in the “new deal for consumers” announced by the Commission and hence, the updating of the consumer acquis.Footnote 116 There is now a Compromise version of the Directive and, according to Article 1 Digital Content Directive (Compromise), the instrument aims to ‘lay down common rules on certain requirements concerning contracts between traders and consumers for the supply of digital content or a digital service’.Footnote 117 More specifically, this provision goes on to note that the Directive aims in particular to establish rules on ‘(1) conformity of digital content/service with the contract; (2) remedies in case of the lack of such conformity or a failure to supply and the modalities for the exercise of those remedies and; (3) modification and termination of such contracts.’Footnote 118As such, the Directive will extend the protections provided to consumers by affording concrete consumer rights and remedies. This is significant as currently at the EU level an infringement of the data protection framework may mean little in terms of consequences for a service contract.Footnote 119 But, despite these good intentions, the Directive raises a number of difficulties from a data protection and privacy perspective. It is important to consider these in detail and in particular, the extension of consumer law to so-called ‘free’ services, as the determination of what will be considered as a price or core term under the scope of the UCT Directive is of key importance in assessing how this Directive may interact with the GDPR. Hence, this Section will first analyze the Digital Content Directive (Compromise) with specific reference to the role of personal data in the various draft versions of the legislation in order to better understand the final Compromise. Building on this, the analysis will then turn to an examination of how the UCT Directive could be interpreted in the context of pre-formulated declarations of data subject consent.
I. Core terms, Passive and Active Collection and Data as Counter-Performance
Article 3(1) of the Commission draft of the Digital Content Directive provided that the proposal “shall apply to any contract where the supplier supplies digital content to the consumer or undertakes to do so and, in exchange, a price is to be paid or the consumer actively provides counter-performance other than money in the form of personal data or any other data.”Footnote 120 Hence, the Commission draft explicitly recognized: (1) the active as opposed to passive supply of data—including personal data—as (2) counter-performance. Both of these points were heavily criticized, in particular by the EDPS’s opinion on the proposal.Footnote 121 The final compromise deletes all explicit references to the active or passive provision of personal data and the term counter-performance. However, the Directive more fundamentally retains the references to the fact that the provision of personal data gives rise to the application of the protections in the Directive. More specifically, Article 3(1) Digital Content Directive (Compromise) provides that:
This Directive shall apply to any contract where the trader supplies or undertakes to supply digital content or a digital service to the consumer and the consumer pays or undertakes to pay a price.
This Directive shall also apply where the trader supplies or undertakes to supply digital content or a digital service to the consumer and the consumer provides or undertakes to provide personal data to the trader, except where the personal data provided by the consumer is exclusively processed by the trader for supplying the digital content or digital service in accordance with this Directive or for the trader to comply with legal requirements to which the trader is subject, and the trader does not process this data for any other purpose.Footnote 122
There are some subtle and arguably significant differences between the compromise version and the draft Article 3(1) in the Commission proposal outlined above. However, the Compromise version of the Directive still raises several difficulties from a data protection and privacy perspective which can be largely placed within two categories general reflecting the points made above in relation to the Commission draft proposal namely: (1) The positioning of personal data as a de facto ‘price’ in a consumer contract and; (2) the delineation of the types of personal data within the Digital Content Directive’s scope of protection. The purpose of this Section therefore, is to analyze these issues in light of the Commission, European Parliament and Council versions and the final Compromise, as both elements present key challenges from a data protection and privacy perspective.
1. Passive and Active Collection and the ePrivacy Directive
As highlighted above, in their draft Digital Content Directive, the Commission confusingly drew a distinction between passive and active personal data collection in Article 3(1) of the proposal. This distinction was further specified in Recital 14 of the Commission draft. From these provisions, it is clear that the intention of the proposal was to exclude personal data such as IP addresses and “other automatically generated information such as information collected and transmitted by cookies, without the consumer actively supplying it, even if the consumer accepts the cookie”Footnote 123 from the scope of application. To add to the confusion, the Commission draft Recital 14 went on to note that the proposed Directive “should also not apply to situations where the consumer is exposed to advertisements exclusively in order to gain access to digital content.”Footnote 124 Although the Parliament amendments proposed the deletion of this distinction,Footnote 125 the Council version retained the separation.Footnote 126 As such, both the Commission proposal and the proposed Council modifications envisaged a distinction between the active and passive provision of data and significantly this has found its way into the Compromise Directive. In this regard it is interesting to refer to Recital 14 Digital Content Directive (Compromise). This provision states inter alia that the Directive should:
[N]ot apply to situations where the trader only collects metadata such as information concerning the consumer’s device or the browsing history, except where this situation is considered a contract under national law. It should also not apply to situations where the consumer, without having concluded a contract with the trader, is exposed to advertisements exclusively in order to gain access to digital content or a digital service. However, Member States should remain free to extend the application of the rules of this Directive to such situations or to otherwise regulate such situations which are excluded from the scope of this Directive.Footnote 127
This Recital therefore, appears to maintain the distinction that was in the Commission and Council versions and this raises two concerns. First, it appears to assume that advertising is separate from any contract formation – albeit while this remains an issue in the competence of the Members States; and second, it draws an odd distinction between different types of personal data. Hence, browsing on sites that do not require log-in information is deemed distinct from visiting a social networking site with only the latter invoking the operation of the compromise Directive. Nevertheless, such an interpretation seemingly disregards the fact that both IP addresses and cookies are widely considered as personal data and that personal data processing for online behavioral advertising purposes requires the consent of the data subject—for example, as confirmed by the Article 29 Working Party in several opinions.Footnote 128
More specifically, although as specified above in Figure 3, contractFootnote 129 and legitimate interestFootnote 130 may be deemed appropriate in a B2 C ISS context under the scope of the GDPR, their applicability to the collection of such passive data provision in the current context is unlikely for two specific reasons. First, within the meaning of the draft versions of the proposalFootnote 131 and Article 3(1) Digital Content Directive Compromise, the provision of personal data necessary for the performance of a contract is excluded from the Directive’s scope of application. Second, in relation to IP addresses, although one cannot deny the potential application of other conditions for lawful processing—such as legitimate interest in Article 6(1)(f) GDPR, as confirmed in the Breyer case—,Footnote 132 in the context of online behavioral advertising, consent is often the condition most likely to be deemed applicable, as the fair balancing act required under Article 6(1)(f) GDPR is unlikely to be satisfied. Of course, this does not comprehensively exclude the potential for advertising that does not require the consent of the data subject—contextual advertising or personal data processing that relies upon the legitimate interest or contract conditions which may relate for example to security purposes or processing necessary for the provision of a service—from being excluded from the scope of the Directive.
Indeed, this point appears to be reflected in the Parliament’s proposed modifications to the Digital Content Directive, given that rather than referring to processing that is strictly necessary for the performance of a contract requested by the consumer—for example, as in the Commission version—reference was made to personal data “exclusively processed by the trader for supplying, maintaining the conformity of or improving this digital content or service.”Footnote 133 Therefore, to summarize the above: The point proposed here is that this draws a line between different types of personal data in a context where the processing of both types requires the consent of the data subject under data protection and privacy law provisions.Footnote 134 Although by removing the terms passive and active, the Compromise Directive aims to avoid the problems associated with delineating based on the manner in which the personal data is provided, the difficulties from a data protection and privacy perspective remain. This criticism is even clearer in the context of cookies.
Malgieri, in his analysis of the Commission proposal, argues that the active-passive distinction is illustrative of an underlying and deliberate legislative intent to develop a personal data taxonomy and to create a separation between “received, observed, inferred and predicted data,” with only received personal data being considered “a legitimate non-monetary payment for the supply of digital content.”Footnote 142 It is submitted here, however, that despite being deliberate, rather than representing an informed legislative choice, the proposal instead manifests the complex legislative history related to the attempted harmonization of contract law formation at the EU level and an apparent disregard for the nuances in data protection and privacy. The EDPS was particularly strong in his criticism of the Commission draft, the EDPS proposes the use of the notion of services as defined in the Treaties to encompass services where no price is paid. To clarify, in the ISS context the e-Commerce Directive includes services financed by advertising,Footnote 143 and the Court of Justice has found that services as defined by Article 57 TFEU do not necessarily require payment by the users.Footnote 144 As a second, but inherently linked alternative, the EDPS suggested that one could refer to Article 3(2)(a) GDPR, which specifies the territorial scope in the GDPR, by stipulating that the GDPR applies where personal data are processed in “the offering of goods or services, irrespective of whether a payment of the data subject is required.”Footnote 145
Nevertheless, despite the EDPS’s suggestions, there is somewhat of a question mark surrounding whether a service, as defined in either the GDPR, e-Commerce Directive, or the TFEU, may be deemed distinct from a service contract.Footnote 146 To clarify, a service contract is defined in Article 2(6) of the Consumer Rights Directive (CR Directive) as “any contract other than a sales contract under which the trader supplies or undertakes to supply a service to the consumer and the consumer pays or undertakes to pay the price thereof.”Footnote 147 In interpreting the scope of the Directive broadly and seemingly in line with Article 57 TFEU, DG Justice has stated in its interpretative guidance of the CR Directive that such contracts do not require the payment of a price by the consumer, but that accessing online services without the express contractual agreement of individuals is excluded from its scope.Footnote 148 Hence, as noted by Helberger, Borgesius, and Reyna, “contracts [for the supply of digital content in exchange of data] that are concluded by tacit agreement would escape the application of the Consumer Rights Directive.”Footnote 149 As pointed to above, this is undoubtedly the root of the confusion and is illustrative of the clear differences between the data protection and privacy, and consumer protection policy agendas. In simple terms, EU consumer protection, as illustrated by the interpretation given to the CR Directive, draws a distinction between express and tacit agreement which simply does not exist in the data protection and privacy framework.
In essence, therefore, it appears that the differentiation between types of personal data or in the Commission proposal—active and passive collection—may stem from the fact that there is uncertainty as to whether a contractual agreement can be formed under all respective Member State contract law traditions in situations where only such passive data are collected. Indeed, from a common law perspective, given the need for consideration for a valid contract formation, it is unclear whether, for instance, browse-wrap contracts necessarily form a valid B2 C consumer contract—unless personal data is considered consideration. As described above, however, one must question how this interpretation aligns with Article 5(3) ePrivacy Directive and the requirement for consent for the storing or accessing of information already stored on the terminal equipment of a user, like cookies, as defined in the GDPR. In simple terms, it is the same consent that will be used to legitimize the provision of both types of personal data, and accordingly, it is difficult to imagine how such a delineation could be justified.
Furthermore, given the intended strength of the definition of consent in the GDPR, it would seem unlikely that this could be positioned as tacit. Indeed, as outlined above in Section B(II), consent as defined in Article 4(11) GDPR requires a “freely given, specific, informed and unambiguous indication of the data subject’s wishes by … a statement or by a clear affirmative action, [signifying] agreement to the processing of [their] personal data.”Footnote 150 Irrespective of such nuancing however, when combined with the conditions for consent in Article 7 GDPR as outlined above, one must question how the delineation between forms of data provision could be deemed in line with the GDPR and ePrivacy Directive requirements.Footnote 151 It is perhaps with this criticism in mind that the Compromise Directive focuses instead on types of personal data. Despite the above the Compromise version de facto retains the distinction and therefore, there is a large degree of uncertainty as to how all this fits together.
2. Counter-Performance or De Facto Counter-Performance
Aside from the confusion surrounding the differentiation between the passive and active personal data provision, albeit undoubtedly connected, there are also clear difficulties with the provision of personal data for access to a service and thus the controversy stemming from the inclusion of the notion of counter-performance in the Commission proposal. In particular, the EDPS highlighted three concerns associated with the use of the term counter-performance. First, the Commission proposal failed to define the term and that the use of one simple catch-all term appears to oversimplify a variety of business models and data usages. Second, linking the active provision of data with the paying of a monetary price is misleading as consumers are often unaware of what they are giving away when it comes to data, and this is not helped by the use of vague and elastic terms to describe the use of the collected data. Third, data and money are clearly not identical, as providing personal data does not deprive an individual of using this same data repeatedly and this complicates matters when it comes to restitution.Footnote 152 Thus, expressly recognizing personal data as counter-performance is controversial from a data protection perspective, because from a normative perspective there are no proprietary rights in personal data.Footnote 153
Although the final Compromise deletes the references to the term “counter-performance,” potential concerns remain, given that the Directive now de facto appears to retain such a role for personal data. The final Compromise stipulates that the Directive applies where the consumer provides or undertakes to provide personal data to the trader. This language appears to have been inspired by the Parliament and Council versions. In particular, the Parliament draft of Article 3(1) stipulated that the Directive:
[S]hall apply to any contract where the trader supplies or undertakes to supply digital content or a digital service to the consumer whether through the payment of a price or under the condition that personal data is provided by the consumer or collected by the trader or a third party in the interest of the trader.Footnote 154
Indeed, at first glance this draft provision seemingly did away with the notion of counter-performance by stating that the proposed Directive applied when access to digital content or a digital service is conditional upon on the provision of personal data. It is argued here however, that the Parliament’s proposed amendments instead merely presented a more subtle recognition of personal data as a de facto quasi-price. This interpretation is further supported when one compares Recital 13 of the Commission draft with the modifications to this provision proposed by the Parliament, see Figure 5 below.
When compared to the Commission version, the Parliament’s proposed modifications of Recital 13 reveal that the omission of the term “counter-performance” in its draft Article 3(1), merely implicitly recognized the same role. In short, the Parliament’s approach instead acknowledged such a status for personal data by recognizing that access to digital content and services can be conditional upon the provision of personal data. This is evidenced by the retention of the term counter-performance in the Parliament draft of Recital 13, as illustrated in Figure 5 above.
The Council’s proposed modifications of Article 3(1) presented another variation, but in essence reflect the underlying intention made apparent from the above discussion of the Parliament amendments. More specifically, the Council version of Article 3(1) stated that:
[T]his Directive shall apply to any contract where the supplier supplies or undertakes to supply digital content or a digital service to the consumer…
It shall not apply … to the supply of digital content or a digital service for which the consumer does not pay or undertake to pay a price and does not provide or undertake to provide personal data to the supplier.
It shall also not apply where personal data are exclusively processed by the supplier for supplying the digital content or digital service, or for the supplier to comply with legal requirements to which the supplier is subject, and the supplier does not process these data otherwise.Footnote 155
As such, the Council modifications replaced the notion of counter-performance and instead framed the role of the provision of personal data in negative terms. More simply, according to the Council version, the Directive does not apply if the consumer does not provide or undertake to provide personal data. This is in contrast to the Commission text that recognized the notion of counter-performance, and the Parliament amendments which provided for situations in which traders give access to digital content or services to consumers either through “the payment of a price or under the condition that personal data is provided.”Footnote 156 Thus, although the Council amendments certainly reflected the discussion above in relation to the acknowledgement of personal data as de facto counter-performance, they also appear to indicative of the Member State reticence in relation to any attempt to harmonize contract law formation at the EU level. Indeed, the Council construction of the provision, in addition to their intention to retain the passive-active distinction, revealed the fear that a positive acknowledgement of personal data as counter-performance in an EU legislative text, even if implicit, would have had an impact on this tightly guarded aspect of national contract law. This is evident in footnote 15 of the Council’s general approach which included the specification that the contractual protections in its version of the proposed Directive can be extended to situations where only passive data are provided where this is considered a contract by national law.Footnote 157 Importantly, the concerns manifested in the Commission and Parliament versions are reflected in the final Compromise.
The Compromise version tries to create a clear delineation between contracts supplied for a price versus those created where the consumer provides personal data. There is some very careful wording in the Compromise in comparison to the Commission proposal incorporating the concerns associated with recognising personal data as an economic asset to be bartered and traded and thus the Council and Parliament versions. Here reference can be made to Recital 13 Digital Content Directive (Compromise) which states inter alia that, “[w]hile fully recognising that the protection of personal data is a fundamental right and therefore personal data cannot be considered as a commodity, this Directive should ensure that consumers are in the context of such business models entitled to contractual remedies.”Footnote 158 In other words, therefore although personal data cannot be considered as a commodity, the provision of personal data can give rise to a contract with the consent of the consumer-data subject—provided the other requirements on contract law formation in national law are met—giving rise to a consumer contract.Footnote 159
For instance, from a common law perspective recognizing that the provision of personal data gives rise to a contract could essentially lead to the personal data being regarded as sufficient consideration.Footnote 160 Therefore, although the use of the term counter-performance attracted much ire, it is perhaps the underlying recognition of the economic value of personal data vis-à-vis contract formation which is the more apt target for such a debate. Somewhat counterintuitively, therefore, data protection enthusiasts critical of positioning personal data as a quasi-price may find support in those critical of contract law formation harmonization. Indeed, as noted by Mak, an important doctrinal question in this regard therefore, is how this provision of personal data will give rise to a contract in national law.Footnote 161 The author goes on to specify that this presents important challenges as inter alia most national contract laws require a monetary payment for a sales/services contract.This matter is far from clear-cut, and although the analysis may seem somewhat tangential to our current focus, the UCT Directive inherently assumes that in the operation of its provisions some price and consideration will be paid—as evidenced by the exclusion of such core terms from the substantive fairness element in the Directive. As a consequence, the determination of what the price is in relation to declarations of consent is an issue worthy of discussion, as it will effectively determine what is exempt. Therefore, the classification of personal data as a price is of key importance to the operation of the UCT Directive, and subsequently, the substantive fairness element contained therein.
II. Data, Price, and Contract Versus Consent
The analysis in the previous sub-Section has revealed a number of questions in relation to the overlaps between contract law and consent, contract, and legitimate interests as conditions for lawful processing in the GDPR. Indeed, given that personal data processing necessary for the provision of the service is deliberately excluded from Digital Content Directive (Compromise), one must question in particular the relationship between consent in Article 6(1)(a) GDPR and contract in a consumer law sense. More specifically, one must question whether consent to personal data processing then necessarily results in a contractual agreement within the operation of the Compromise Directive. With this in mind, how the overlaps between contract as a condition for lawful processing in Article 6(1)(b) GDPR and consent in Article 6(1)(a) GDPR, as well as how the UCT Directive, and more specifically, the protection against unfair pre-formulated declarations of data subject consent fit within this complex interwoven legal framework can also be questioned.
1. Necessity and the Role of Consent
Building on the above, it is important to remember Article 7(2) GDPR laying down the requirement that consent must be presented in a manner which is clearly distinguishable from the other matters, and Article 4(11) GDPR which stipulates that consent must be freely given. The requirement for the separation of consent and other matters is reinforced in Article 7(4) GDPR which states that:
[W]hen assessing whether consent is freely given, utmost account shall be taken of whether, inter alia, the performance of a contract, including the provision of a service, is conditional on consent to the processing of personal data that is not necessary for the performance of that contract.Footnote 162 [Emphasis added]
As noted by the EDPS, this separation is further evident in Recital 43 GDPR which provides that where “the performance of a contract, including the provision of a service, is dependent on… consent despite such consent not being necessary for such performance,” there is a presumption that consent is not freely given.Footnote 163
In analyzing these provisions, one must wonder what this separation of consent and other matters—for example, the details of the contract—means from a theoretical perspective regarding the classification of the pre-formulated declaration of consent subject to the contractual protections afforded by the UCT Directive. If consent is to be separated from the provision of the service, how can the GDPR rely on the application of the protections against unfair terms in pre-formulated declarations of consent in its recitals if personal data is to be viewed as the price for the provision of the service? In other words, as consent precisely constitutes a lawful condition for the processing of an individual’s personal data and the personal data would at the same time constitute the de facto price for the provision of the service to the individual in a B2 C consumer contract sense, it seems at first glance to be counterintuitive to present consent in a manner which is clearly distinguishable from the other matters as required by Article 7(2) GDPR. This challenge manifests itself even more clearly when one recognizes the distinction between consent and contract as conditions for lawful processing.
Article 6(1)(b) GDPR provides that personal data processing is lawful where such “processing is necessary for the performance of a contract to which the data subject is party or in order to take steps at the request of the data subject prior to entering into a contract.”Footnote 164 Importantly, the notion of necessity has its own distinct meaning within the scope of the GDPR. In essence, this refers to the necessity component in the fair balancing element of the fairness principle in data protection. The question thus becomes what is necessary for the performance of a contract and hence, the argumentation around whether—and, if so, which—personal data processing operations are integral to the delivery of free services and the economic underpinnings of the internet. The Article 29 Working Party has repeatedly noted that it seems unlikely that large scale personal data processing for commercial purposes—for example, online behavioral advertising—would satisfy this necessity test.Footnote 165 There is therefore a distinction between consent and contract as conditions for lawful processing but what does this mean in terms of the conditionality of consent and thus the relationship between consent and consumer?
Here it is important to note that the Working Party has explicitly stated “the GDPR ensures that the processing of personal data for which consent is sought cannot become directly or indirectly the counter-performance of a contract.”Footnote 166 Despite the fact that this is an ongoing debate, from a systematic reading of the GDPR it is indeed difficult to position personal data as a price. Such doubt also appears to reflect a teleological interpretation of the GDPR in that although the Regulation has dualist aims—for example, as it also focuses on the integration of the internal market—it is more predominantly weighted towards the protection of fundamental rights and freedoms and in particular, the right to data protection—as evidenced by its reliance on Article 16 TFEU as its legislative basis. Indeed, if one was to position personal data as the core term, how would this align to the fair balancing, and hence, the proportionality and necessity components? More specifically, the very purpose of Article 4(2) UCT Directive is to leave such matters for the contractual parties. Although data protection is not an absolute right, it nevertheless establishes key fairness checks and balances in the GDPR which must be respected.
Nevertheless, as the recent policy initiatives show, this matter is far from simple to understand. In this regard one must wonder how the Article 29 Working Party opinion on consent aligns for instance with the modifications of the consumer law acquis. How can the Article 29 Working Party issue an opinion that appears to contradict the Digital Content Directive Compromise and the new deal for consumers? And, therefore, should the freely given stipulation be understood not as a strict requirement, but instead merely as an indicator that utmost account shall be taken of whether access to the service is conditional on consent? An important point of reference here is the recent opinion by Advocate General Szpunar in the Planet 49 case where he notes that “[…] from the terms ‘utmost account shall be taken of’, the prohibition on bundling is not absolute in nature.”Footnote 167 There is clearly a large degree of uncertainty here as to how these provisions and frameworks are to be interpreted. Hence, it remains to be seen not only how far consent will be stretched, but also how processing that is necessary for the contract will be delineated from additional activities. The Digital Content Directive as a legislative development, provided it is formally published in the Official Journal, should therefore be considered to have more authority than the non-legally binding opinions of the Article 29 Working Party. Ultimately, the boundaries of the notion of consent will need to be determined by the Court of Justice.Footnote 168 Hence, the classification of what constitutes the core terms for the purposes of the UCT Directive remains uncertain, and this is indicative of the teething problems inherent to the alignment of the data protection and consumer protection policy agendas. This uncertainty illustrates the fundamental divide between the dominant view in data protection, which positions personal data protection as a fundamental right, and the approach in consumer protection and competition law which are increasingly recognizing and catering to the economic value of personal data.Footnote 169 Although the ability to offer higher levels of protection also extends to the exclusion of the exemption provided in Article 4(2) UCT Directive in the national implementation,Footnote 170 legal certainty and a fully harmonized approach appear to be at risk. In this regard it is important to remember that Recital 42 GDPR refers to pre-formulated declarations of consent specifically.
2. Positioning Consent and Applying Contractual Protections
Although it is clear from the above analysis that consent and contract are separate and entirely distinct conditions for lawful processing in the GDPR, this separation raises some doubt as to the positioning of consent in relation to the contractual protections in the UCT Directive. In other words, can consent to a pre-formulated declaration be understood as a contract in its own right despite its required separation from the provision of a service contract? Can consent in Article 6(1)(a) GDPR itself be reduced to a form of contract? And can consent to a pre-formulated declaration legitimizing personal data processing in effect act as a trigger for the formation of a B2 C consumer contract?
This interpretation does not, however, render contract and consent synonymous—as contract law assumes the autonomous decision making capacity of individuals. Therefore, although the formation of a contract requires the voluntary assent of the parties, consent in data protection cannot be reduced to a form of contract, given that it must not always be freely given, specific, informed, and unambiguous as understood under the GDPR, in order for it to be considered a B2 C contract. This is indicative of the fact that the UCT Directive focuses on the fairness of the terms themselves and explicitly excludes the analysis of the validity of the contract formation. As such, the validity of a data subject’s consent and the fairness of the pre-formulated declaration of consent are two connected, but distinct, issues. Nevertheless, this higher threshold for consent in data protection does not exclude the possibility that the data subject’s consent may give rise to a B2 C contract. Indeed, this currently hinges on whether the provision of personal data can be recognized as conditional for the provision of the service in national contract law, or indeed on whether national contract law otherwise recognizes the existence of a B2 C contract.Footnote 173 The Article 29 Working Party opinion should therefore be taken with a grain of salt, as this is an issue which is far from resolved. As such, much hinges on the interpretation of the Digital Content Directive (Compromise) and Article 7 GDPR by the Court of Justice, as well as the reform of the ePrivacy Directive and therefore the proposed ePrivacy Regulation.
From the above, therefore, although the legislator intended to cross-reference the protections in the UCT Directive in Recital 42 GDPR for pre-formulated declarations of consent, there are many unresolved issues centered around the recognition of the economic value of personal data. Despite the fact that the EDPS and Article 29 Working Party have criticized the positioning of personal data as a form of payment, this remains a highly contentious issue with divergences in interpretation amongst policy makers, academics, and even different enforcement bodies.Footnote 180 Indeed, at first glance, it seems unlikely from a systematic and teleological interpretation of the GDPR that personal data could be positioned as the price, given the separation of data subject consent from other matters in order to be certain that it will not fall foul of the freely given requirement; despite this fact, this position has not been reflected in consumer protection and competition law and policy. The uncertainty regarding the recognition of the economic value of personal data described above raises a number of issues in terms of how the GDPR may be interpreted in practice and by the Court, in particular in light of the Digital Content Directive (Compromise). One must therefore wonder whether the Article 29 Working Party opinion on consent is truly sustainable in the current regulatory environment. The action taken by Max Schrems referred to in the introduction will hopefully provide the answer to this question, however, one must wonder whether a more tiered understanding of conditionality is in fact needed.
E. Data Protection and Freely Given Consent—A Framework Designed to Counteract Imbalances?
I. Take It or Leave It Choices and Freely Given Consent
Where consumers are confronted with take-it-or-leave-it offers and do not have a real choice but to accept the terms and conditions if they want to use a particular service, it seems difficult to ensure that consent is freely given. It is seemingly on this basis that the Article 29 Working Party has drawn a strict line dividing processing necessary for the provision of a service, and other processing requiring consent, in its interpretation of Article 7(4) and Recital 43 GDPR. Due to the existence of economic characteristics such as network effects, economies of scale and economies of scope, the markets in which online businesses compete are typically characterized by the presence of only a few firms that have a rather large market share. Individual control over personal data is becoming illusory when dominant companies are able to impose their practices on individuals by exploiting their strong position. This may result in an imbalance of power between individuals and providers of online services, which calls into question the existence of a genuine choice for data subjects as to whether or not to give their consent to a particular form of personal data processing.Footnote 182 As such, one can question the appropriateness of consent as a condition for lawful processing where only a limited number of providers are present in the market or where one provider is dominant.
Recital 43 GDPR makes these issues explicit and provides that consent should not be a valid legal ground where there is a clear imbalance between the data subject and the controller. Although the recital continues by referring, in particular, to the situation where a public authority acts as a controller where “it is therefore unlikely that consent was freely given in all the circumstances of that specific situation,”Footnote 183 the described imbalance is not exclusive to situations in which public authorities act as controllers, but is also applicable in the reality of current concentrated online markets.Footnote 184 This interpretation of Recital 43 GDPR is supported by both the Article 29 Working Party and the EDPS,Footnote 185 and consequently, there appear to be two components to consider in the assessment of the freely given stipulation in this context, namely: (1) An assessment of all the circumstances of that specific situation which is activated only if (2) a clear imbalance exists between the controller and the data subject. The GDPR, therefore, appears to require an assessment of the controller-data subject asymmetry in abstract—for example, in general and not with specific reference to the particular context of the data subject(s)—in order to establish if a clear imbalance exists. Subsequently, an analysis of the established imbalance is needed in order to assess whether the freely given requirement has been violated, or whether it is justifiable in the circumstances of that specific situation. Nevertheless, it is difficult to interpret what these elements may incorporate concretely.
It is suggested here that in keeping with the accountability principle, the controller may be required to prove not only that informed, specific, and unambiguous consent has been provided in line with the requirements in the GDPR, but also that the clear imbalance in power did not affect the consumer-citizen’s decision to consent, despite the fact that this consent was required to access the service in question. It is not at all clear what this may mean, given that the GDPR works from the assumption of an asymmetrical controller-data subject relationship and that hence, it is hard to imagine a situation where there would not be an asymmetric controller-data subject relationship, especially in an ISS context. This echoes the point above regarding imbalance and the UCT Directive in Section C(II)(1). With this in mind, one must question what is to be understood by a clear imbalance in practice or in other words, the circumstances in which this presumption of the failure to satisfy the freely given consent stipulation would be rebuttable. Interestingly, in its opinion on consent, the Article 29 Working Party interprets the provisions as precluding a controller from arguing that the data subject’s consent was freely given on the ground there is a choice between its services and those of competitors.Footnote 186 The Working Party notes that such an interpretation would render the data subject’s freedom of choice dependent on: (1) Other market players; (2) whether the data subject actually deemed the services equivalent; and would also (3) require the controller to constantly monitor competitors to ensure the validity of the data subject’s consent. This would clearly have an impact on legal certainty.
Nevertheless, given the above policy developments and the uncertainty surrounding the future of the proposed ePrivacy Regulation and indeed the seemingly imminent adoption of the Digital Content Directive (Compromise), it seems that the Working Party’s approach may need to be revisited in the future. Indeed, there is an ongoing fundamental debate surrounding the merits of surveillance capitalismFootnote 187 and the ongoing legitimacy of the monetization of personal data, as evidenced by the contrast between the reforms and the Working Party’s opinion on consent. Do we want to ban business models centered around the monetization of personal data? Or force companies to offer alternative personal data-based and pay-for-access, monetary funded versions of the same service? Personalized advertising is certainly not the only way of monetizing online services. But would the second option not de facto also put a price on the rights to data protection and privacy? These are all fundamental and normatively challenging questions which need to be answered. It seems unlikely, however, that the Article 29 Working Party’s strict interpretation of consent, and its separation of processing necessary for the provision of the service will be sustainable in light of the various moves to recognize the economic value of personal data and the broader internal market considerations of the EU legislator.
Moreover, although there is strong merit to each of the Working Party’s arguments listed above, it is suggested that the only reasonable consequence of failing to engage with the question of the controller’s position on the market would be to acknowledge the conditionality-availability distinction between consent and explicit consent implicit. More specifically, if the proposed ePrivacy Regulation is adopted and in keeping with the Digital Content Directive (Compromise) directly or indirectly allows for the rendering of consent to be conditional for access to services, by continuing to refuse to factor in market power in the assessment of the freely given stipulation in the future, the Working Party could conceivably be left with for instance, the delineation of consent and explicit consent as the sole means of assessing conditionality. One might wonder whether such an approach would respect the spirit of the GDPR and the risk-based approach inherent to the interpretation of the Regulation’s requirements. In short, the potential legislative recognition of the legitimacy of conditional access based on consent may force the Working Party to reconsider whether Facebook’s and other market players’ position should be a consideration in the assessment of the validity of consent under the freely given stipulation. Indeed, in this regard, one must question whether smaller market operators should be denied rendering consent conditional where they are merely a minor player on the market. In these circumstances, there is less need for a strict interpretation of consent, as the risk that there is no free choice for data subjects is countered by the presence of other market players to which data subjects can switch. A strict approach may even discourage small businesses and start-ups from entering the market, thereby reducing the choice that the Article 29 Working Party in fact aims to protect. Consequently, is this the desirable choice, or could there not be a role for competition law in identifying situations in which the conditionality of access would have an impact on data subject consent?
Data protection advocates have expressed increasing attention in competition enforcement in recent years.Footnote 188 These debates have so far mainly addressed the question of how data protection interests can be considered in the competition analysis, and how competition enforcement may thereby strengthen the effectiveness of data protection law. But, the complementarity of the two regimes could also work the other way around.Footnote 189 The use of competition principles in data protection law seems particularly promising in the interpretation of the scale of the obligations with which controllers and processors have to be in compliance. Competition concepts of market definition and dominance could play a useful role here; the stronger the position of a controller or processor in the market, the riskier the processing activities for the right to data protection of the individual. It is worth noting in this regard that the GDPR does consider the level of risk of a certain form of processing in such a way that more detailed obligations will apply to controllers where the risk of processing is higher. For example, the risks of varying likelihood and severity for the rights and freedoms of natural persons play a role in determining to what extent the controller must implement appropriate technical and organizational measures to ensure, and be able to demonstrate, that the processing of personal data is performed in compliance with the applicable rules under Article 24(1) GDPR. In a similar vein, the Court of Justice referred to the ubiquity of online search engines in Google Spain when determining the effect of the interference caused by Google’s processing of personal data with the fundamental rights to privacy and data protection of the data subject at issue.Footnote 190
Unlike competition law, data protection law is not concerned with scale because a breach of data protection rules can be equally damaging to the interests of individual data subjects irrespective of the market position of the firm and the size of the dataset or the processing activities.Footnote 191 Nevertheless, while no formal distinction is made on the basis of scale or size under EU data protection law, the risk inherent in particular processing activities, and the ubiquitous nature of a controller, can thus be considered as relevant factors in establishing, respectively, the scale of its obligations under the GDPR and the impact of its processing activities on the rights of the data subject. These factors resemble, at least to a certain extent, the well-established principles of market definition and dominance in competition law. In this regard one can refer to a number of examples throughout GDPR in order to highlight the prioritization of risk and, thus, the risk-based regulatory model employed in the Regulation.
More specifically, aside from the higher risks and stricter requirements associated with the processing of sensitive personal data and, thus, delineations based on data type, the GDPR also refers to the scale of the processing operations—for instance, in the requirements relating to the appointment of a data protection officer and the exercising of a data protection impact assessment. In short, these provisions reflect the role of the fair balancing element of the GDPR’s fairness principle and, thus, the application of the proportionality and necessity principles as components of this element. As a context dependent assessment, competition law analysis could provide valuable insights into the practical application of the fair balancing element via the principles of market definition and dominance. Such considerations can be used to examine the availability and viability of alternative services, and thereby the extent of citizen-consumer decision-making capacity in particular circumstances.
Competition law reasoning could, therefore, be used to interpret key data protection concepts such as fairness and accountability. In this sense, the stronger the position of the controller and, thus, the less chance for data subjects to rely on another controller, the stricter the principles of fairness and accountability would need to be applied to adequately protect the interests of data subjects.Footnote 192 With regard to the interpretation of the concept of consent, this logic would imply that the existence of dominance in a competition law sense may act as an indicator challenging the validity of consent as a condition for the processing of personal data. In this vein, one could question whether controllers will be able to differentiate between the services that they offer and hence, whether premium services could be offered to those willing to pay either a monetary fee or with their personal data, or even with both. As a result, one must question how such a requirement would map against existing practice and hence, the likelihood of a strict interpretation. Indeed, in this regard one may wonder how this interpretation would affect the business models of companies such as Google and Facebook, given their strong market position in search, social media and also online advertising. It is therefore uncertain what role the freely given stipulation will play in terms of the initial citizen-consumer sign up and thus the requirement to offer the service with personal data processing limited to only that which is necessary for the performance of the contract for the provision of the service.
Interestingly, the German competition authority, Bundeskartellamt, opened an investigation against Facebook in March, 2016,Footnote 193 which specifically targets the interaction between market dominance under competition law and the validity of consent under data protection law. In February, 2019, the Bundeskartellamt concluded that Facebook abused its dominant position in the market for social networks by infringing data protection rules. In particular, the Bundeskartellamt prohibited Facebook from combining user data from different sources, except if users voluntarily consent to their data being combined. According to the Bundeskartellamt, Facebook’s terms and conditions violate data protection law and thereby also constitute exploitative business terms under the abuse of dominance prohibition of competition law. According to the Bundeskartellamt, it cannot be assumed that users effectively consent to Facebook’s collection and use of data from third-party sources in view of its dominant position on the market. In the words of the president of the Bundeskartellamt:
As a result of the decision of the Bundeskartellamt, Facebook has to restrict its collection and combining of data. Assigning data of Facebook-owned services like WhatsApp and Instagram as well as of third party websites to Facebook user accounts is only possible subject to the voluntary consent of users. If consent is not given, the data has to stay with the respective service and cannot be processed in combination with Facebook data. It is up to Facebook to develop proposals to implement the limitations imposed. With regard to Facebook’s future data processing policy, the president of the Bundeskartellamt stated that “we are carrying out what can be seen as an internal divestiture of Facebook’s data.”Footnote 195 And that:
In future, consumers can prevent Facebook from unrestrictedly collecting and using their data. The previous practice of combining all data in a Facebook user account, practically without any restriction, will now be subject to the voluntary consent given by the users. Voluntary consent means that the use of Facebook’s services must not be subject to the users’ consent to their data being collected and combined in this way. If users do not consent, Facebook may not exclude them from its services and must refrain from collecting and merging data from different sources.Footnote 196
While the Bundeskartellamt is relying on competition enforcement to address unfair collection and use of personal data by incorporating data protection principles into competition law, its investigation also illustrates the relevance of competition law principles for defining the validity of consent within data protection law. In particular, the reasoning of the Bundeskartellamt may be interpreted as an acknowledgement that the existence of dominance in a competition law sense points to a clear imbalance between the data subject and the controller, resulting in a rebuttable presumption that the data subject’s consent has not been freely given.Footnote 197 Considering the restrictive nature of the concept of dominance, however, this competition law principle should not be the sole indicator.
II. Freely Given and the Data Necessary for the Provision of a Service
Dominance is defined in case law of the Court of Justice as “a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by giving it the power to behave to an appreciable extent independently of its competitors, customers and ultimately of its consumers.”Footnote 198 Very large market shares are in themselves, and save in exceptional circumstances, evidence of the existence of a dominant position.Footnote 199 Other factors that are taken into account in the assessment of dominance include the existence of entry barriers that make it difficult for other companies to access the market. The concept of dominance is thus rather restrictive considering that a company will only be found dominant when it can behave independently on the market. As a result, a clear imbalance for the purposes of the GDPR may arguably still be present even if no dominance can be established from a competition law perspective. The fact that a company is not dominant should, therefore, not automatically result in the conclusion that there can be no clear imbalance in the relationships with its users. For instance, the presence on the market of three undertakings holding market shares of about thirty percent each will normally preclude any finding of dominance,Footnote 200 while such a market structure may still not provide effective alternatives to users if the three undertakings, together capturing ninety percent of the market, each have similar privacy policies in place. Although the presence of dominance in competition law terms arguably obviates the need for an additional consideration of whether there is a clear imbalance, its absence should not preclude such a finding altogether.
Aside from this more macro level market dominance-based argumentation regarding the existence of an imbalance, it is important to reiterate that asymmetries are inherent to the operation and underlying presumption of data protection law. Accordingly, it must be understood that even when there appears to be no clear imbalance between the parties, the validity of the data subject consent may still be questioned and, thus, as reflected in the GDPR, the conditions mentioned in the definition of consent—namely that consent must be freely given, specific, informed, and unambiguous—still need to be satisfied. Hence, although the determination of what a clear imbalance means in practice may incorporate and benefit from a competition law analysis, even in the absence of a clear imbalance the freely given requirement still needs to be satisfied. This is manifested and further reinforced by the burden of proof on controllers to be able to demonstrate valid data subject consent in Article 7(1) GDPR.
The freely given stipulation clearly goes beyond an analysis of market dominance and the availability of choice, even in the above interpretation, but also incorporates an analysis of the ability of the data subject to choose freely—for example, freedom from inter alia any form of intimidation, coercion and or deception.Footnote 201 The assessment of the freely given stipulation is, therefore, not only strongly context dependent, but is also tied to the other conditions—specific, informed, and unambiguous—mentioned in the definition of consent.Footnote 202 For instance, in interpreting what is meant by the term unambiguous, it is useful to refer to the requirement for controllers to be able to demonstrate that the data subject has in fact consented to the processing of their personal data as mentioned above, for example, Article 7(1) GDPR. Additionally, and through a combined reading of Articles 7(1) and 7(4) GDPR—thus, in light of the freely given stipulation—an unambiguous indication would also seemingly require that any such consent is demonstrably connected to the stated specific purpose and that the data subject is informed prior to the giving of the consent and, thus, any resulting processing, and that this prior information is presented in a “distinguishable … intelligible and easily accessible form, using clear and plain language”.Footnote 203 Indeed, in this regard one can refer to the Article 29 Working Party opinion on consent which clearly specifies that blanket consent which fails to indicate the scope and consequences of the processing clearly would not be considered specific.Footnote 204 This reflects the interwoven nature of the conditions and is illustrative of the importance of the transparency principle in the GDPR and its substantive aspect, and is further supplemented by the more formal requirements for prior information in Articles 13 and 14 GDPR in terms of the information to be provided to the data subject before the processing of personal data. In addition to these ex ante requirements, one can also refer to the Section 1 Transparency and Modalities in Chapter 3 Rights of the Data Subject of the Regulation.Footnote 205 Article 12 GDPR relating to transparent information, communication, and modalities for the exercise of the rights of the data subject, implicitly provides the same distinct manifestations of the transparency principle.Footnote 206 As such, the transparency principle manifests both substantive and formal aspects in both an ex ante and ex post context.
But what does all this effectively mean in practice due to the fallibility of consent? Indeed, given the well-documented failures of consent, one must question how these conditions will be interpreted in practice and how effective they will be despite the changes in the GDPR.Footnote 207 To reiterate, it remains to be seen how far consent will stretch, but also if, and how, processing that is necessary for the performance of a contract will be delineated from additional activities as required by Article 7 GDPR, or if in practice personal data will be recognized as a de facto price or counter-performance. In practice, it is arguable that the role for the freely given stipulation may most readily find its application in updates to existing terms. Indeed, such a prediction is hardly surprising, and in this regard one can again refer to the fallout from the recent Facebook-WhatsApp merger. Such an assessment, as reflected in the AGCM rulings, may be less of an analysis of the fairness of terms under the UCT Directive, and instead may invoke comparisons with the application of the Unfair Commercial Practices Directive (UCP Directive) vis-à-vis the means through which consent was attained.
Although a thorough examination of the UCP Directive remains outside the scope of this Article, it should be noted that distinguishing where data protection enforcement begins, and consumer protection ends—or vice versa—as well as delineating what warrants a contractual or practice orientated fairness assessment, is highly debatable. It is clear, however, that changes in privacy policies are, and will continue to be, viewed skeptically regarding the validity of data subject consent. Irrespective of the specific issues at the core of that particular case, however, it will also be interesting to see if a similar reasoning may be applied for more general updates without the additional merger element. In this context, one should remember the fall-out from Google’s decision to merge its privacy policies for its various services in 2012. But, given the ongoing discussion in the proposed ePrivacy Regulation, the issues remain largely up in the air. This uncertainty is particularly prevalent in the interpretation of when a clear imbalance will occur and thus how the applicability and operation of this rebuttable presumption will work in practice.
Pre-formulated declarations of consent should respect the data subject interests and, therefore, follow the reasonable expectation that these texts should be drafted with the balancing of interests in mind. In simple terms, privacy policies and pre-formulated declarations of consent should live up to their name and, in essence, respect the data protection fairness principle. This view of the role of pre-formulated terms is far from the practical reality, which raises clear concerns regarding the protection of the right to data protection and its underlying rationales of autonomy and informational self-determination as protected in the GDPR in an effort to tackle informational power asymmetries.Footnote 208 In response to this divergence between the law on the books and the law in practice, there has been an increasing use of the UCT Directive in the assessment of terms relating to the processing of personal data. This is also reflected in the GDPR, with Recital 42 GDPR specifically referencing the UCT Directive regarding pre-formulated declarations of consent. The precise contours of the relationship between the GDPR and UCT Directive are uncertain given that, first, the GDPR is an omnibus regime, whereas the UCT Directive refers specifically to B2 C contractual agreements only—for example, consent in the GDPR applies to much more than just B2 C contexts—, and, second, the UCT Directive represents a far more economic assessment as opposed to the fundamental rights approach evidenced in the GDPR.
Despite these differences, however, Benöhr contends more generally speaking that the consumer protection agenda may be furthered by a broad range of Charter rights, including for instance the right to data protection.Footnote 209 This appears consistent with the aim of integrating consumer interests in all relevant policies and the goal of targeting more systematic consumer protection as expressed in the European Consumer Agenda in 2012.Footnote 210 As such, it is arguable that the adoption of the Charter may breathe new life into consumer protection policy.Footnote 211 That said, there are major issues which need to be ironed out.Footnote 212 As illustrated in this Article, key challenges await from both a perspective of consumer contract formation as well as a data protection and privacy view. These difficulties are most aptly illustrated in the debates surrounding the recognition of personal data’s economic value, and this is clearly a polarizing debate. The legislator seems set on ploughing ahead with maximum harmonization law-making, however. Indeed, due to its apparent failure to approach such issues with true regard for all relevant policy agendas, it is unclear how the Digital Content Directive (Compromise) will be interpreted following its publication in the official journal and the passing of the 2 year transition period, and what this will mean for the data protection and privacy framework. Irrespective of such concerns, however, there is a willingness in consumer protection and competition law circles to recognize the economic value of personal data in contrast to the opinion of the Article 29 Working Party.
That said, the data protection and privacy community is increasingly turning towards consumer protection and competition law to help provide more holistic citizen-consumer protection. These respective policy agendas will have to meet each other in the middle somewhere, and it may be left to the Court of Justice to sort out the conceptual mess. The analysis in this Article has shown that in the context of pre-formulated declarations of consent, such a collaborative role can be carved out with respect to the relevant frameworks and the intent of the three respective policy agendas. A more concurrent substantive application of protections with delineated but complementary enforcement of the respective frameworks is needed to empower citizen-consumers.