Digitized Terms: The Regulation of Standard Contract Terms in the Digital Age

2.1 Platform (Dis-)Intermediation

In supply chains, digital transformation is replacing traditional intermediaries with new forms of distribution: Direct, peer-to-peer contracts between suppliers and buyers are becoming more common, with intermediation instead taking place on online platforms. Unlike the intermediaries who used to take ownership of the traded goods, these digital marketplaces do not have to become parties to the supply contract themselves, but merely mediate these contracts.^[12] The so-called platform economy and the sharing economy, based on online platforms such as AirBnB, BlaBlaCar and Uber, are key features of the digital economy.^[13] Due to their diversity, online platforms are difficult to define. Some authors describe them as marketplaces where providers offer intermediation services based on data305 processing,^[14] while others focus on the Internet as a technical connection.^[15] By creating digital structures that allow two groups of users (suppliers and customers) to meet and transact with each other, online platforms form a central pillar of digital markets. They enable consumers to find online information and allow businesses to exploit the advantages of e-commerce.^[16] However, they raise widely discussed competition and private law questions, for instance about their responsibility in relation to illegal or harmful content or products hosted in the frame of their operation.^[17] As a centerpiece of Europe’s digital strategy, the Digital Markets Act (DMA) has recently been adopted.^[18] It aims to ensure fair behavior by large online platforms that act as gatekeepers in digital markets.^[19] In addition, the Regulation on Platform-to-Business relations (P2B Regulation) has been enacted in 2019 in order to ensure the transparency of, and trust in, the online platform economy in business-to-business relations.^[20]

From the perspective of the UCTD, the characteristic feature of online platforms is that transactions are governed by the respective platforms’ terms, in addition to – or even in place of – State-made contract law. Platforms provide the framework ‘that shapes the way in which supply and demand are matched, in which products are perceived, and hence in which a vast number of individual contracts are formed’.^[21] Platforms typically operate as marketplaces with a triangular structure where users306 must first conclude a user contract with the platform to be subsequently able to conclude contracts between themselves.^[22] As part of this user contract, they agree to the platform’s terms and conditions.^[23] Their regulatory content, however, typically extends beyond the bilateral relationship between the user and the platform and has a profound impact on the individual transactions between the users.^[24] By this means, online platforms largely substitute the conventional default rules of contract law with their own, private rules: In fact, they act ‘as de facto “lawmakers” for the platform environment’.^[25] The so-called platform law is developing into an autonomous normative order of digital platforms, based in particular on standard contract terms.^[26]

As opposed to the traditional mode of application of standard terms, these indirect effects of online platform terms do not fit well with the rules of the UCTD as they do not form part of ‘contracts concluded between a seller or supplier and a consumer’ (Article 1 para 1): To market partners, the platform is a third party.^[27] Platforms also call the applicability of other consumer law rules into question: As opposed to bank lending, for instance, peer-to-peer lending is not subject to regulatory safeguards such as bank deposit guarantee schemes, and it is unclear whether respective lenders are protected by the rules on consumer credit contracts.^[28] Generally speaking, the emergence of the platform economy modifies the regulatory framework within which transactions occur. Online platforms substitute their own platform terms for the otherwise applicable default rules, thereby affecting the sophisticated governance mechanism of307 knowledge production characterized by the interaction between public and private rule-making.^[29] Platforms substitute the conventional legal framework of private autonomy by their own, private rules, at least in part.^[30]

2.2 Contractual Personalisation

In addition to enabling online platforms, digital technologies provide tools for handling large amounts of data. They enable data analytics aimed at discovering useful information, drawing conclusions, and supporting decision-making by transforming and modelling data on a large scale.^[31] The two phenomena are linked because the underlying data often originate from platforms. It may include preferences of market participants, for example for contract terms. Artificial intelligence and machine learning enable dynamic adjustments based on data about supply and demand, competitor activity, and delivery schedules.^[32] While practices of algorithmic pricing are among the most widely discussed manifestations of this technological development,^[33] the same technological means also allow algorithmic adjustments to other contractual clauses, in particular dynamic adaptions of platform terms. Online platforms are therefore in a position to use data analytics to dynamically adjust or even personalize their terms and conditions. However, personalized contractual terms do not fit easily into the regulatory scheme of the UCTD. It does not even seem clear whether they have been ‘drafted in advance’ within the meaning of Article 3 para 2 UCTD or whether they fall outside the scope of the Directive. There has been little discussion of this issue so far: Even though platforms have much better access to the necessary data than legislators, at least in the field of contract law, the focus of the legal debate has been on personalized law rather than on personalized terms.^[34] 308

In any event, algorithmic capabilities to personalize interactions with consumers to exploit information asymmetries and consumer biases raise fundamental private autonomy concerns.^[35] The digital technologies of data analytics and artificial intelligence enable algorithmic agents to replace human choice in contractual transactions and decisions.^[36] But if technological devices have an impact on contractual decisions, private autonomy may be at risk. The challenge for any regulatory assessment is to understand the differences between human and machine decision-making in order to assess the impact on private autonomy.^[37] Rulemakers are considering regulatory strategies to avoid discriminatory or unfair algorithmic decisions, notably in the context of the European AI Act Proposal.^[38] However, if algorithmic data analysis is redefining the way markets operate, it seems necessary to reflect these new risks of unfair contract terms in its specific regulation, rather than leaving them (largely) outside its scope.

2.3 Self-Enforcement

Digital technologies also provide alternatives to legal enforcement by enabling private parties to impose or exclude certain actions by technological means. The use of data and connected devices can be restricted, for example by preventing smart locks from opening if rent is not paid. As a consequence, the technological infrastructure replaces legal mechanisms, and a ‘lex cryptographica’ emerges.^[39] Smart technologies operate as a mechanism of automatic enforcement that functionally resembles and partially replaces judicial enforcement. In particular, blockchain and distributed ledger technologies allow for the design of so-called smart contracts: Smart contracts are self-executing agreements written in code on the blockchain.^[40] Because such agreements can be automated by computers and are enforceable through the309 tamper-proof execution of software code, they take on tasks traditionally assigned to legal institutions.^[41] While distributed ledger technologies offer alternative, digital mechanisms for contract enforcement, their relationship to the legal sphere is still unclear.^[42] If the enforceability of (personalised) platform terms is ensured on such a technological basis, the legal regulation of unfair contract terms runs the risk of being rendered meaningless. At the very least, self-execution shifts the burden of litigation to the other party.^[43] In a recent, controversial decision the German Federal Court of Justice ruled that a standard contract term was invalid because it allowed the lessor of a car battery to remotely block the recharging option after the extraordinary termination of the rental agreement: According to the court, the fact that the lessee can only continue to use the battery (and thus the separately purchased electric car) after successfully claiming the transfer of use in court creates a significant imbalance.^[44] Technology-based self-enforcement thus places the UCTD in a delicate tension with smart contracts. On the one hand, its fairness control could significantly limit the legal admissibility of smart contracts, but on the other hand, these legal limits threaten to become irrelevant as smart contracts make their enforceability more burdensome. The recently adopted ELI Principles on Blockchain Technology, Smart Contracts and Consumer Protection attempt to strike a balance by postulating in Principle 18 lit d that in the case of self-enforceable unfair terms, consumers shall be entitled to immediate redress by way of re-coding of the smart contract.^[45] What the Principles do not explain, however, is how such a remedy could effectively be enforced. The explanatory notes state somewhat apodictically that ‘any problems arising as a consequence of this will have to be solved by the applicable law in light of the acquis communautaire’.^[46] 310

3.1 Transparency Obligations

A key regulatory tool of standard term regulation is the requirement of transparency. Article 5 UCTD provides that standard terms shall ‘always be drafted in plain, intelligible language’. While the maxim that ‘sunlight is the best disinfectant’ forms the basis of market-based regulation in general,^[49] the so-called information model based on transparency is paradigmatic for large parts of European private law.^[50] The potential – and limitations – of this information model have long been at the heart of the debate on the regulation of standard terms.

The European legislator has adopted the same regulatory approach in the broader context of digital legislation: the P2B Regulation refers to transparency in its title and makes it one of its main regulatory objectives (cf Article 1 para 1). In particular, its Article 3 requires providers of online intermediation services to ensure transparency in various different respects. However, it is doubtful whether311 the regulatory instrument of transparency will be sufficient to preserve private autonomy in the face of digitized standard terms.^[51] While transparency rules can help to overcome information asymmetries and promote informed choices, they prove ineffective when restrictions on private autonomy have other causes. In particular, such rules cannot counteract the increasingly coercive nature of platform conditions resulting from network effects, behavioral correlations, and self-enforcing mechanisms.^[52] They can therefore only serve as an effective regulatory tool if there is either a choice between different platform terms or, alternatively, a chance to negotiate different terms. Otherwise, transparency rules will not be sufficient to achieve the regulatory objective, but can at best be used as an additional instrument to complement broader regulatory strategies. After all, an informed decision requires the existence of choice in the first place.

3.2 Fairness Review

Another, more intrusive regulatory tool to deal with standard terms is the judicial control of their fairness. According to Article 3 para 1 UCTD ‘a contractual term which has not been individually negotiated shall 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’. Article 6 para 1 UCTD provides that such term shall not be binding.

This regulatory strategy is also being applied in the broader digital context. Like transparency, fairness is one of the regulatory objectives of the P2B Regulation (cf Article 1 para 1) and is also mentioned in its title. Article 8 explicitly refers to the standards of good faith and fair dealing, but only in relation to very specific contractual terms and situations (such as retroactive changes to terms and conditions). The broader provision on standard terms in Article 3, on the other hand, does not explicitly refer to fairness in its para 1 lit a, but instead focuses on transparency. The regulatory instrument of fairness review is therefore not very strongly articulated in the P2B regulation, probably because in most jurisdictions, judicial review of standard terms is considered to be justified only in B2C relations. Exceptionally, however, there may be ‘a case for a special control of P2B terms and conditions based on the established rationale for controlling general terms and conditions in contract312 law’ even in the B2B context, namely where a platform has become a gatekeeper.^[53] Some authors, therefore, argue in favor of judicial fairness review in the context of the P2B Regulation, despite the absence of an explicit provision.^[54]

In contrast to the P2B Regulation, Article 13 of the proposed Data Act (DA Proposal) explicitly provides for a fairness test.^[55] The rule is aimed at contractual terms that concern ‘the access to and use of data or the liability and remedies for the breach or the termination of data related obligations’. It does not only protect consumers but also includes standard terms that are imposed on (at least certain) enterprises. Similar to the regulatory approach of the UCTD, unfair contractual terms are declared non-binding (cf Article 13 para 1 DA Proposal). Moreover, the Data Act does not simply refer to the general principles of good faith and fair dealing, but specifies fairness by means of a black and grey list in paras 3 and 4, again similar to the UCTD.^[56] In contrast to that Directive, however, the more general yardstick for fairness is ‘good commercial practice in data access and use’ (cf Article 13 para 2 DA Proposal), rather than legal default rules.^[57] The subject matter of data access and use is simply not yet sufficiently covered by such rules: ‘Up to now, neither model terms nor default rules suitable for the implementation of mandatory access rights are sufficiently developed’.^[58]

In this way, the provision illustrates a much more fundamental problem for the fairness control of digitized standard terms in general: Whereas default rules typically serve as a yardstick for assessing deviating standard terms,^[59] their relevance decreases significantly with the increasing prevalence of ‘platform law’. Taken together, online platforms, supported by algorithmic data analysis and self-executing313 technologies, have the potential to replace State-created default rules by platform terms.^[60] This tendency could be further accelerated by other contemporary developments, some of which are related to digital transformation. For instance, the heteronomous efficacy of default rules is more limited in transnational private law transactions which are at least facilitated by digital technologies, in particular by globally operating online platforms.^[61] In addition, technological innovations may have the effect that the regulatory models provided by the legislator no longer reflect the changed reality or are no longer fit for purpose because they do not address certain issues raised by these innovations.^[62] Such innovations are often driven by digital transformation. Until recently, for example, the law of sales did not offer any suitable default rules for payment with data.^[63]

Contrary to what Article 13 DA Proposal seems to suggest, the regulatory solution cannot be to replace default rules with good commercial practices across the board. In a digital environment, such practices can only develop in the face of the structural risks of digital transformation outlined above. Because of these risks, emerging commercial practices will not be able to provide a neutral, balanced benchmark against which terms can be effectively measured. In addition, self-enforcing technologies reduce judicial control over standard terms; by providing alternative mechanisms for enforcing platform terms, these technologies reduce the need to go to court.^[64] Both because of the impending lack of a fairness standard and because of these enforcement deficits, fairness review does not promise to be an appropriate regulatory tool for dealing with digitized standard terms.314

3.3 Opt-Out Restrictions

An even more intrusive regulatory approach would be to generally restrict deviations from default rules, irrespective of the substantive content (and, in particular, the fairness) of the standard term in question. Such a strategy of ‘regulating opt-out’^[65] amounts to a tightening of the altering rules, in particular by requiring that alterations be based on an explicit agreement between the parties to the exchange contract. This would exclude any spill-over of platform terms into the contractual relationships of the platform users. In a (roughly) comparable manner, Article 5 UCTD prohibits deviations by standard terms that are not drafted in plain, intelligible language: Unintelligible terms cannot alter default rules, regardless of their substantive content. A similar ‘clarity-requiring altering rule’^[66] could be applied to trilateral situations on digital platforms, based on the argument that alterations are not sufficiently intelligible when provided for in platform terms. The alterations, it could be argued, are not sufficiently intelligible to the two parties when they bilaterally agree on the exchange contract.

Such a restriction would effectively prohibit platform law. Platform terms would no longer be able to supplement the exchange contracts concluded on that platform. However, a complete exclusion of the contractual freedom of platform operators seems questionable against the background of the constitutional guarantee of private autonomy.^[67] It would not only affect significantly unbalanced platform terms but would also restrict impartial platform operators.^[68] It could even hamper the platform economy, in particular with regard to transactions for which legal default rules do not provide an adequate solution. Filling such gaps with good commercial practices could have a counterproductive regulatory effect if such practices are themselves dictated by platform terms. In any case, European law does not provide for such far-reaching restrictions. Instead, the P2B Regulation confines itself to requiring that platform terms be made available at all times (cf Article 3 para 1 lit b):^[69] Instead of restricting altering rules, the rule-maker opted for a simple requirement of transparency of platform terms.315

4.1 Limiting Conflicts of Interests

The regulatory approach of limiting the regulatory competence of platforms subject to a conflict of interest faces the difficulty of concretizing this vague requirement. On the one hand, market power could be used as an indicator, but is itself difficult to determine. Even the rules on standard terms in the P2B Regulation are not linked to this criterion, although concerns about the dependence on large, powerful platforms clearly influenced their adoption.^[76] Instead, the regulation applies across the board to all companies, regardless of any market power threshold.^[77] On the other hand, conflicts of interest are particularly likely to arise on so-called dual-role platforms where the platform operators themselves – or persons associated with them – are also participants on that platform.^[78] If platform operators themselves act as sellers on the platform (along with independent sellers), it is obvious that they have an interest in formulating terms that are favourable for sellers rather than for buyers.^[79] Such dual-role platforms are self-interested when defining the platform law.^[80] They cannot be considered to be impartial, even with regard to those platform transactions where only independent parties interact. Due to the potential conflict of interest between the roles of rule-maker and market participant, neutral and balanced platform conditions are unlikely to emerge. Market participants cannot be expected to trust these terms.

A variety of legal rules on conflicts of interest can potentially serve as a model for regulation. For instance, Article 8 of the European Crowdfunding Service Providers for Business Regulation (ECSPR) provides rules on conflicts of interests of providers who operate or manage digital platforms in order to provide crowdfunding services by matching business funding interests of investors and project owners (cf Article 2 para 1 lit a, d, and e ECSPR). The various requirements that Article 8 ECSPR imposes on crowdfunding service providers include a ban on participating in crowdfunding offers on their platform (para 1), restrictions on project ownership by related parties (para 2), the obligation to maintain and operate effective internal rules to prevent conflicts of interest (para 3 and 4), and various disclosure obligations (para 4–6).^[81] While this317 provision is primarily concerned with the operation of a platform rather than with its rule-making, it nevertheless illustrates the general principle that platform operators subject to conflicts of interest are subject to legal constraints. Similar restrictions should apply with regard to standard terms,^[82] for example by declaring them non-binding if provided by a conflicted platform operator, or at least by excluding the spill-over effects of platform terms on the exchange contracts concluded on the platform. Whereas such restrictions would reach too far if they applied as a general rule, they seem entirely appropriate under the condition of a conflict of interest.

4.2 Promoting Trustworthiness

Restrictions on conflicted platform operators may help to prevent particularly dubious platform terms that are likely to be unfair, but they do not promote good commercial practice. By declaring platform terms to be non-binding, they even create contractual gaps that need to be filled in the absence of appropriate default rules. To promote gap-filling rules, an additional regulatory mechanism is needed that encourages the emergence of balanced terms. Rather than sanctioning conflicted rule-makers, such a regulatory instrument should privilege rule-makers who appear to be particularly trustworthy. Such an approach would take advantage of the fact that platforms are, in principle, particularly well placed to design rules that mimic the market, as they have access to data on market participants’ preferences that is usually unavailable to the legislator.^[83] Default rules have traditionally provided a sophisticated mechanism for managing knowledge production, but they are becoming less functional as a result of the current digital transformation.^[84] A regulatory instrument that promotes private regulatory knowledge could therefore prove beneficial in restoring the interaction between public and private rule-making, in order to recreate the legal infrastructure necessary for the well-functioning of private autonomy.

However, it is challenging to design provisions that privilege trustworthy rule makers, both in terms of their scope of application and the nature of the privilege. The Data Governance Act (DGA) provides some regulatory guidance.^[85] Its rules on318 data altruism aim to promote specific, particularly trustworthy intermediaries.^[86] Even if the provisions of Article 15 et seq. DGA concern data intermediaries and not platforms as rule makers, knowledge production is a common denominator: Both are characterized by the use of data, and the common regulatory challenge is the need to define criteria for distinguishing and privileging intermediaries that can be considered particularly trustworthy. In order to qualify as a trustworthy organization, Article 16 DGA requires, in particular, that legal entities be established to achieve objectives of general interest and to operate on a non-profit basis.^[87] Similar criteria, albeit with some adjustments, could be applied to platforms. Economic incentives are the main driver of significant imbalances in platform terms. If platforms aim to be particularly attractive to one of the two sides of the market, their terms tend to be unbalanced. When profit orientation is excluded, this incentive is removed. To ensure that the actual operation of the platform remains commercially attractive, the rules could, for example, allow the task of rule-making to be outsourced to independent not-for-profit bodies. In addition to the requirement to meet general interest objectives, dominant platforms should be excluded from the regulatory privileges: Market dominance, even if it is not based on a profit motive, severely restricts the regulatory choices available to platform users.^[88]

The DGA also provides much less guidance in terms of the regulatory privileges that can be granted. While the DGA grants them only in relation to data access and collection, the privileges of trusted platforms should aim at strengthening their rules. On this basis, balanced terms could evolve into a market standard of good commercial practice. For instance, the terms of trusted rule-makers could benefit from an exemption from fairness review under the UCTD. While the Directive was originally tailored to asymmetric relationships between the two contracting parties (cf recital 2 UCTD), such a relationship does not exist in trilateral platform constellations, after all. Significant imbalances may still occur, but they are unlikely to evolve if the platform is trustworthy in accordance with the mentioned criteria. In the absence of incentives to discriminate between platform users, the likelihood of fair platform conditions is good enough to exempt them from the UCTD. Such an exemption would, in turn, encourage platform terms that are better suited than traditional default rules to fill the gaps in incomplete platform contracts. Digital technologies give platform operators better access to the data needed to reflect319 parties’ preferences and ‘mimic efficient arrangements’^[89] than the legislator itself could possibly have.^[90]