Competition and consumer policy in digital markets

2. The intersection between competition and consumer policy in digital markets

As digital markets have reshaped the way consumers interact with firms and the way firms compete, they have created new challenges for regulatory and enforcement frameworks. Regulators and policymakers are increasingly confronted with practices that may have implications for both consumer and competition law. In light of these developments, this section aims at examining how these two frameworks interact and is divided into two parts. It first outlines the foundational objectives of competition and consumer policy, examining how the distinctive characteristics of digital markets have challenged the ability of each framework to fulfil its goals. It then considers different types of conduct that highlight the increasing convergence between the two policy domains in the digital environment. While the section does not aim to provide an exhaustive list of practices relevant to both fields, it identifies several prominent areas that illustrate their growing interdependence.

2.1. The policy goals

Before delving into the intersection between consumer and competition policy, it is useful to briefly outline their core objectives, and note how these objectives have evolved in response to digitalisation, an evolution well-documented in past OECD work (e.g. (OECD, 2023[21]; 2022[15]; 2020[16]; 2010[22]; 2008[1])).

Competition policy aims to safeguard the effective functioning of markets. When markets function well, consumers benefit from a wide range of goods and services at prices shaped by competitive rivalry. Competition intervention focusses primarily on addressing market failures arising from market power (ICN, 2022[23]).1 Because its objective is to preserve or restore the competitive process, competition policy adopts a broad understanding of markets that extends beyond the parameters of individual transactions, incorporating considerations on market structure and conduct (Okiche and Okiche, 2025[24]). It also adopts a wide notion of consumer. While it considers final consumers’ welfare, it may also recognise businesses as (intermediate) consumers when they act as customers that buy goods or services and, in that role, are affected by market conduct (Nathani and Akman, 2017[25]).

Competition policy relies on supply-side instruments to achieve its goals. This includes prohibitions on cartels and other anti-competitive agreements, abuses of dominant position or attempted monopolisation, and, in general, practices that distort the competitive process of markets. Unlike competition policy that tackles failures that are external to consumers, consumer policy directly targets consumers to empower them to make well-informed decisions, and to protect them from misleading, fraudulent and unfair commercial practices, and from unsafe products (Graef, 2021[26]; European Commission, 2025[27]).2 For this reason, the concept of “consumer” is narrower than in competition policy and refers specifically to individuals purchasing goods or services for personal, non‑business purposes.

While competition and consumer policy across jurisdictions may have other objectives and goals,3 both policies share a common ultimate purpose of safeguarding consumer welfare.4 They are aimed at fostering efficient markets that benefit consumers and businesses alike. Synergies between the two policies have also widely recognised transparency of information at the core of their interaction. Rules that ensure accurate and transparent product and transaction information not only safeguard consumers but also enhance competition by encouraging firms to compete on their merits (Tonazzi, 2025[2]). To the extent that consumers are placed in a better position to consciously and actively exercise their choices, consumer protection rules can also be seen as having a pro-competition role (ICN, 2022[23]). This means that, together, competition and consumer policy create an environment in which businesses can compete on a level playing field and consumers can make informed choices (UNCTAD, 2025[28]). The complementary nature, as well as the existence of overlapping concerns in the two policy frameworks persist in digital markets. Robust consumer protection and empowerment can lead consumers in online environments to make better choices, thus playing an important role in driving competition and innovation. Likewise, more competition in digital markets can lead to better business practices, contributing to consumer empowerment and an overall increase in consumer welfare.

Product safety provides a particularly clear illustration of the complementarity between competition and consumer policy in digital markets, especially in the context of online marketplaces. While product safety is primarily a matter of consumer protection, it also affects competition because it forms part of the quality of goods and services on which firms may compete. Where businesses reduce costs by cutting corners on product safety standards, businesses that invest in compliance, testing and safer design may be placed at a competitive disadvantage (Australian Treasury, 2019[29]). This risk can be amplified in online marketplaces, where the scale, speed and cross-border reach of online distribution can allow unsafe products to reach large numbers of consumers before authorities are able to intervene. In this way, effective product safety enforcement can serve both consumer and competition policy objectives by helping to ensure that competitive advantage is not derived from non-compliance or the externalisation of product safety risks onto consumers.5 Moreover, the proliferation of unsafe products online, as identified in a 2021 OECD sweep report, may erode consumer trust in online marketplaces and in online transactions more broadly, to the detriment of compliant sellers as well as consumers (OECD, 2023[30]).

These complementarities do not, however, eliminate the possibility of tension between the two policy areas in particular cases. For example, competition enforcement might inadvertently reduce consumers’ ability to make informed choices if the resulting changes in business strategies fail to account for information asymmetries or consumer biases. This may occur, for instance, when competition remedies involve changes to choice architecture that generate limited impact or unintended consequences arising from the existence of consumers’ default biases (Fletcher and Vasas, 2024[31]). Concerns have also been raised in the opposite direction. While stronger product safety-related obligations on online marketplaces and traders can protect consumers, they could potentially increase compliance costs and raise barriers to entry for smaller firms or new market entrants. Such requirements may, in some circumstances, disproportionately benefit large incumbents, who are better equipped to comply, while placing a heavier burden on smaller firms with limited resources. Like with other types of regulation, smaller sellers may face higher administrative and financial burdens, slowing their growth and limiting competition, while large incumbents are better positioned to absorb these costs (ICN, 2022[23]).6

2.2. Consumer agency in digital markets

In digital markets, the question of consumer agency, that is, the extent to which individuals can understand their options, act on their preferences, and meaningfully influence market dynamics, has become increasingly central (Wertenbroch et al., 2020[32]). Agency has never been static, but digitalisation has introduced layers of intermediation and design choices that subtly reshape how consumers navigate markets. As online environments have grown more complex, the tools available to firms to structure choice, personalise content, and adapt interfaces in real time have become markedly more sophisticated (Zac et al., 2025[33]). This evolution has expanded opportunities for innovation and efficiency, yet it has also made it easier for design features to influence decisions in ways that consumers may not fully perceive (US FTC, 2022[34]).

Rather than resulting from overtly misleading practices alone, the erosion of consumer agency, also understood as consumer autonomy, often materialises through incremental design decisions (small, iterative changes to the design): how options are framed, how information is sequenced, or how friction is distributed across a user journey (Davida, 2024[35]). These elements are central components of online choice architecture (see ‎2.3) and their impact on consumer agency underscores their relevance from the perspective of the two policies: they are the means by which interface design affects both consumer outcomes and competitive dynamics. These mechanisms do not remove choice, but they can affect how choices are interpreted or pursued, particularly in environments where attention is limited and cognitive demands are high.

Understanding the pressures on consumer agency therefore requires looking beyond single design tactics and instead considering a broader set of market dynamics.7 Modern digital interfaces are frequently optimised through continuous testing and large-scale behavioural data, enabling firms to refine how consumers encounter information or alternatives (Fast, Schnurr and Wohlfarth, 2023[36]). These optimisations may genuinely enhance usability, but they can also tilt decision making towards outcomes that align with firms’ commercial incentives. Where such dynamics are present, consumers may find it more difficult to compare alternatives, reassess their preferences, or switch to competing services.

In digital markets, agency can be affected when the presentation of information obscures relevant details or emphasises some choices disproportionately (Kozyreva, Lewandowsky and Hertwig, 2020[37]; Islam, Ali and Azizzadeh, 2024[38]). Also, the growing cognitive complexity of digital environments means that even small changes in framing or defaults can meaningfully influence behaviour (Jin, Zhang and Chen, 2017[39]). Moreover, the practical ability to switch providers may be weakened by design choices that complicate cancellation, limit interoperability, or discourage multi-homing (Siciliani and Giovannetti, 2019[40]). Finally, data portability and interoperability increasingly determine whether consumers can carry their preferences, histories, or connections with them, affecting whether switching is not only desirable but feasible (Pecher, Syrmoudis and Grossklags, 2024[41]).

These pressures do not operate in isolation. When agency is weakened, consumer protection concerns, such as transparency and informed choice, intersect with competition concerns regarding entry, switching and effective rivalry (Graef, Clifford and Valcke, 2018[42]). In other words, design choices that limit the meaningful exercise of choice can also diminish the competitive constraints faced by firms. This convergence has led to a growing recognition that consumer and competition policy cannot be treated as strictly separate domains in digital markets. Box 1 presents some considerations on how the most recent developments in AI, particularly related to AI agents, may impact consumer agency. This issue is explored further in Section ‎3, which examines how the two policy frameworks could contribute to each other in practice and how authorities have begun to operationalise it.

Box 1. Consumer agency in a world with agentic AI

As advances in artificial intelligence and the growing reliance on autonomous AI agents and tools continue to reshape market dynamics, both consumer and competition policy are confronted with emerging tasks and novel challenges.

Agentic AI, AI systems that autonomously take actions on behalf of humans, can already compare products, negotiate prices and even place orders. This means that they can interact in digital environments and take actions on behalf of consumers according to pre-defined preferences and payment options. These new dynamics will reshape markets, and their effects could span consumer and competition policies as well as other areas such as privacy, all of which are human-centric.

Literature has already started looking at possible effects, for instance, on consumer agency, and has signalled that agentic AI can both expand and constrain consumer autonomy depending on design, governance and external market dynamics. The level of impact depends on how much delegation there is to the AI agent, reshaping in any case what consumer agency means and how it can be taken into account in consumer and competition enforcement. Consumers may choose how much control to give to the agency, shifting autonomy from low delegation to full decision making.

On the positive side, AI agents can empower consumers to delegate tasks such as research, comparison and purchasing without involving biases inherent in humans. They can also facilitate consumers express their preferences and articulate their needs to better match their demand to the available supply. This could also lead to lower search, comparison and shopping costs. All these processes can also happen at unprecedented speed, contributing to faster and more optimal decisions, greatly amplifying consumer agency.

When transactions become AI-driven, however, there is also a risk of reduced consumer agency that comes from the agent predetermining what the consumers see and are able to explore. Consumers may become less aware of product choice or brand identity as they rely more on agentic filtering. This risks shifting consumer autonomy from final consumers to AI agents, who may act as intermediaries, potentially creating new gatekeepers. If the AI agent is wrongly trained, errors in its output can also distort consumer autonomy, nudging its choices in unintended directions. This may intensify if there is overreliance or minimal oversight on the AI agent.

From a competition perspective, agentic AI also raises distinct concerns. Integrated platforms that simultaneously operate marketplace infrastructures and deploy AI agents face inherent conflicts of interest, creating incentives for self-preferencing in how agents rank, recommend or transact. More broadly, the emergence of agentic AI may entrench the position of platforms already dominant in search and data, given that the effectiveness of an AI agent depends critically on access to large datasets, reinforcing existing ecosystemic advantages.

As agentic AI becomes more embedded in market interactions, the already complex relationships between consumers and businesses are set to evolve in new and potentially far‑reaching ways. Advancing research on the impact of AI agents, especially on the nature and exercise of consumer agency, will therefore be crucial in guiding future policy development.

Sources: OECD (2025[8]), Artificial intelligence and competitive dynamics in downstream markets, https://doi.org/10.1787/ccf0624a-en; Busch, C. (2025[43]), Consumer Law for AI Agents. Elsevier BV, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5187056; Hunt S., et al, (2025[44]) Will 2025 be the year of the agent? A primer for competition practitioners on the next wave of AI innovation. Competition Law & Policy Debate, 9(1), 20-30, https://www.elgaronline.com/view/journals/clpd/aop/article-10.4337-clpd.2025.0004/article-10.4337-clpd.2025.0004.pdf.

2.3. Conduct at the intersection of the two policy areas

The second part of this section turns to specific types of conduct in digital markets that may trigger intervention. It examines practices that can fall under consumer policy, competition policy, or simultaneously within both frameworks. The discussion highlights complementarities between the two approaches as well as potential areas where their respective powers and responsibilities may overlap, signalling how co-ordinated or complementary enforcement can be crucial in addressing complex digital‑market behaviours (as will be further discussed in Section 3).

To provide a structured frame for the analysis that follows, the paper will discuss the following types of behaviour where the overlap between the two policies is particularly evident: personalisation, misleading or deceptive online choice architecture and a set of practices with foreclosure effects. For each conduct, it first presents the key concerns raised from each policy perspective, outlining possible effects and enforcement alternatives under each framework. Then, it identifies aspects that bridge both policy areas and sets the stage for the discussion in Section 3 on joint approaches.

2.3.1. Personalisation practices

Personalisation (mostly through prices but also through other variables such as advertising and recommendations) is the “practice of (price) discriminating final consumers based on their personal characteristics and conduct” (OECD, 2018[17]). This results in consumers acquiring a product tailored to their preferences, for example at prices that reflect closely their individual willingness to pay. In digital markets, where businesses are able to collect higher amounts of consumers’ data, price personalisation has become a reality (BEUC, 2023[45]), with firms increasingly able to tailor offers in ways that approach the theoretical ideal of first-price discrimination in microeconomic theory (see Box 2 for key definitions). While the discussion focusses mostly on price personalisation, it acknowledges the existence of other personalisation practices including rankings and targeted advertising8 and their analogous effects on consumers and competition.

Previous work, including past OECD discussions, has explored the benefits of personalisation, which include online experiences that are tailored to consumers’ needs and preferences, with increased engagement and availability of goods and services for consumers that would have otherwise not been able to consume9 (see for example (OECD, 2023[12]; 2023[46]; 2018[17])). From a mere economic perspective, the closer the discrimination to perfect discrimination, the higher the allocative efficiency (i.e. the better the alignment between consumers and producers’ preferences), with the consumer welfare being transferred to the firms (Varian, 1989[47]) as explained in (OECD, 2018[17]). In imperfect personalisation scenarios, the resulting levels of efficiency are lower, with some transactions that could have been efficient not happening (for consumers with a higher willingness to pay), generating what literature knows as deadweight welfare loss and an overall decrease in output (Motta, 2004, p. 496[48]).

Box 2. Key definitions in personalisation practices

Price discrimination

In economics, there are traditionally three degrees of price discrimination being the first degree the scenario where perfect discrimination among customers materialises, with each consumer being charged a price equal to its willingness to pay. Second degree discrimination refers to situations in which businesses grant discounts once a specific purchase quota is achieved, with prices being set in two parts. This means that they offer the same product at different prices for different quantity or quality levels. Third degree price discrimination takes place when businesses charge different prices to different groups of customers based on certain characteristics that groups them.

In digital markets, there are common practices that account for price discrimination at different degrees, depending on the level of big data analytics performed over consumers’ information (Ezrachi and Stucke, 2016[49]). Examples are steering and re-offers. Steering, also known as search discrimination, refers to differentiation by search engines of consumers even when they submit the same search query. Re-offers are the practice of presenting an offer at a discounted rate to a consumer that has searched for a specific product or service, giving an advantage to “more patient” consumers.

Other personalisation practices

While price is the most used variable to discriminate among customers, digital businesses can also discriminate through other variables, such as advertising and rankings. Ads personalisation means adjusting ads to match individuals’ interests and characteristics, while personalised rankings involve practices that adjust order of appearance of products or services in search results. In both cases, businesses use information about the user beyond the search query (including previous queries, location, browsing history and purchasing behaviour) to decide which results to display and in what order.

Sources: Varian, H. (1989[47]), Chapter 10 Price discrimination. In Handbook of Industrial Organization, Handbook of Industrial Organization Volume 1 (pp. 597-654). Elsevier, https://www.sciencedirect.com/science/chapter/handbook/pii/S1573448X89010137; Ezrachi, A., & Stucke, M. (2016[49]), Virtual Competition. Journal of European Competition Law & Practice, 7(9), 585-586, https://academic.oup.com/jeclap/article/7/9/585/2547746?guestAccessKey=; OECD (2018[17]), Personalised Pricing in the Digital Era, https://www.oecd.org/en/publications/personalised-pricing-in-the-digital-era_db4d9c9c-en.html.

Despite these benefits, concerns regarding the use of personalisation practices have been raised from both competition and consumer perspectives. While they are not prohibited per se under either of the two regimes, they may account for infringements under certain circumstances (see (OECD, 2023[14]; 2025[50])).10

From a consumer policy perspective, personalisation may raise concerns where it is opaque, misleading, exploitative or combined with other practices that impair consumers’ ability to make informed choices (OECD, 2023[46]). In other words, under consumer law the issue is not personalisation as such, but whether the way it is designed, presented or implemented gives rise to a misleading, unfair or otherwise harmful commercial practice. A central concern is information asymmetry, which is exacerbated in digital markets: consumers will often not know how prices, offers, rankings, recommendations or advertising have been tailored to them, making it difficult to assess available options, compare alternatives, or detect differential treatment (Bergemann and Bonatti, 2024[51]). This informational advantage may allow businesses to exploit not only behavioural biases common across consumers, but also the situational vulnerabilities of individual consumers (OECD, 2023[46]).11 Consumer detriment may take multiple forms, including financial, privacy and psychological impacts, and is likely to be greater for consumers who are less discerning, less engaged or who face higher switching costs (OECD, 2023[46]). More broadly, the prevalence of personalisation in digital environments may require reconsideration of how concepts such as the "average" or "reasonable" consumer are applied in enforcement, since personalisation means that each consumer may face a different offer, price or presentation, making it difficult to assess a commercial practice against a single representative benchmark (OECD, 2023[46]).

From a competition framework, when personalised practices for consumers are implemented by a dominant firm, it may result in exploitative effects towards consumers and, under certain circumstances, exclusion of competitors from the market. While both theories of harm have been analysed by market studies and reports conducted by competition authorities in the past years, enforcement in this area, particularly in digital markets, remains limited.

In multiple jurisdictions exploitative abuses are recognised as abusive conduct, in principle. However, in others, mainly in common law jurisdictions, there is a traditional rejection of purely exploitative conduct as antitrust violations. When exploitative abuses fall within the scope of competition enforcement, personalisation can be conceived under the rationale that some consumers can be charged higher prices for reasons not related to costs (i.e., their higher willingness to pay for the goods or services). In digital markets, where “winner-takes-most” dynamics prevail, this may imply higher barriers to entry, thus leaving consumers more locked in without a choice but to accept the unfair terms or excessive prices for a relevant period of time. This means that personalisation may result in excessive pricing, or discriminatory or unfair terms, thus having the potential to constitute exploitative abuses. In practice, however, enforcement priorities in most jurisdictions have traditionally focussed on exclusionary abuses, leading to less clear guidance on how authorities could potentially deal with the exploitative effects of personalisation in digital markets (Balasingham and D’Amico, 2024[52]; OECD, 2018[17]).12 It is worth noting that exploitative conduct in digital markets is not confined to personalisation. While personalisation is examined here as a prominent example, exploitative practices may occur independently of personalised treatment, and in practice there is often a thin line between exploitative conduct, exclusionary abuse and self-preferencing. These categories, though analytically distinct, may describe overlapping aspects of the same conduct. The appropriate enforcement framing may depend on legal and strategic considerations rather than on the nature of the behaviour itself.

Personalisation can also result in the exclusion of rivals from the market. This may happen because rivals may not be able to match the asymmetric access to granular consumer data that allows the dominant firm to customise its offers, reinforcing barriers to entry and, thus, impacting the competitive process, as a whole. When these strategies entail selectively offering lower prices to a competitor’s customers in a way that could hinder the competitor’s ability to compete, this may give rise to concerns about potential exclusionary conduct (ICN, 2022[23]).

Some challenges arise when analysing personalisation under abuse of dominance or monopolisation provisions. Competition authorities or competent bodies typically bear the burden of establishing when personalisation amounts to anti-competitive conduct, including by demonstrating either a competitive disadvantage for affected customers or its capability to distort competition.13 The difficulty in proving, under the competition framework, whether terms or price are unfair or excessive is higher in digital markets, where often pricing reflects the multiple sides of the markets, making pro-competitive effects harder to distinguish from anti-competitive conduct (Botta and Wiedemann, 2019[53]). Moreover, because in digital markets prices constantly change over time and adjust to new information and data received, the extent of their excessiveness or unfairness will also vary, making more difficult for the competition authority to conduct its analysis (Graef, 2021[26]). When personalisation occurs for reasons linked to heterogeneity in costs (e.g., search costs), there may be objective justifications that restrict the application of competition provisions (UNCTAD, 2021[5]). One final challenge relates to whether the good or service in itself was personalised, as this would make more difficult to argue that transactions between the different types of transactions were equivalent (Botta and Wiedemann, 2019[53]).

Overall, when assessing personalisation practices, competition and consumer regimes converge in important ways. Competition law asks whether they reinforce a dominant position or entrench market power, thus hindering effective rivalry, whereas consumer law evaluates whether they meet the standards of transparency. This overlap between both areas, as will be analysed in Section 3 below, offers multiple enforcement routes to regulators examining personalised practices. At the same time, the analytical challenges associated with evaluating personalisation under competition law may give consumer protection enforcement a more prominent role.14 This is especially true where unfair practices provisions are available, as their assessment follows a more normative approach (i.e., relies on clearer standards rather than on market effects). Remedial strategies to resolve concerns brought by personalisation are also at the intersection of both policies. Requiring businesses to disclose the use of personalisation practices (information duties) may be available in both enforcement areas and could encourage consumers to shop around.15

One final important element that is worth considering is the behavioural relevance of personalisation for consumers. Behavioural studies seem to suggest that consumers consider “unfair” the lack of transparency in online markets, rather than price discrimination per se.16 This further blurs the need for intervention from a competition perspective, at least against personalisation itself, revealing a preference for transparency and opt-out remedies rather than explicit prohibitions (see Section 3).

2.3.2. Misleading or deceptive online choice architecture practices

Online choice architecture “is the environment in which users act, including the presentation and placement of choices and the design of interfaces” (CMA, 2022[54]). In well-designed choice architectures, consumers may face themselves with seamless purchase and after-sales processes, relevant recommendations and opportunities for future actions. However, certain choice architecture practices may also impact consumers negatively when they are deceptive, misleading, fraudulent or otherwise illegal (OECD, Forthcoming[55]). These may include practices involving framing, creating a sense of urgency, generating social proof, forcing registration or information disclosure, nagging to make a choice, or making it difficult to cancel or opt out (OECD, 2022[3]) (see examples in Box 3). These practices make user’s behaviour feel organic, thus appearing as an exercise of free will (Day and Stemler, 2019[56]).

Box 3. Examples of misleading or deceptive online choice architecture practices

The OECD presents a useful taxonomy of misleading or deceptive online choice architecture practices (2022[3]). The identified categories are:

• Forced action: practices that aim at forcing consumers to do something in order to access a specific functionality. Examples include forcing them to register or into disclosing more personal information than required, including consumer’s contacts.

• Interface interference: practices that aim at exploiting framing or anchoring biases by privileging specific actions for consumers through framing of information. Examples include preselection of options favourable to the business by default; giving visual preference to some options, thus creating a false hierarchy; displaying misleading reference prices; using trick questions such as double negatives; and framing through emotive language.

• Nagging: practices that involve repeated requests to consumers to do something favourable to the business. This includes turning on notifications or location-tracking.

• Obstruction: practices that aim to make a task flow or interaction more difficult than needed with the intent to dissuade an action. Examples include making it easy to sign up for a service but hard to cancel it or opt out of certain settings. Similarly, practices like generating click fatigue to steer consumers to choose the simpler option or preventing price comparison.

• Sneaking: practices that seek to hide, disguise or delay information relevant for the consumer’s decision. Examples include non-optional charges that are added later to the final price (known as drip pricing), sneaking an item into a consumer’s basket without previous consent and automatically reviewing a transaction following a trial period (hidden subscriptions).

• Social proof: practices that involve proof attempt to trigger a decision. For example, notifications about other consumers’ activities and purchases or testimonials (that may be misleading or false).

• Urgency: practices that impose a real or fake sense of urgency to pressure consumers into making a purchase. Examples are messages of low stock or high demand, or countdown timers to indicate an expiring deal or discount.

Source: OECD (2022[3]), “Dark commercial patterns", OECD Digital Economy Papers, No. 336, OECD Publishing, Paris, https://doi.org/10.1787/44f5e846-en.

Misleading commercial practices are, of course, not a product of digitalisation: evidence of consumers being deceived by traders can be traced to antiquity. What technology has changed is the scale, precision and speed at which such practices can be deployed, enabling real-time behavioural optimisation across millions of users simultaneously, a dimension with no historical precedent. However, in online environments, choice architecture practices have become prevalent.17 This happens because businesses in digital markets have more possibilities to run repeated experiments with consumers that allow them to steer them into making choices that may not be in their best interests (OECD, 2023[12]). For example, the use of algorithms enables businesses to use consumers’ data at a speed and scale that allows them to test their choice architecture in real time. This possibility increases firms’ awareness of how behavioural insights can be leveraged to refine their marketing strategies, including in ways that exploit consumer biases (Narayanan et al., 2020[57]).

From a consumer policy perspective, these practices, whether related to structure, information or pressure,18 and regardless of market power, may then have the effect of misleading, deceiving, or coercing consumers, causing direct or indirect detriment to consumers (CMA, 2022[54]).19 In addition to impairing consumers’ autonomy, some of these practices may cause significant financial loss, privacy harm and psychological detriment. Thus, they may be prohibited under consumer law against deceptive or unfair business practices.20

While there is evidence of recent enforcement action from consumer protection authorities against different types of misleading or deceptive online choice architecture practices,21 their prevalence and opacity could make it difficult for enforcers to identify the most severe and recurrent offenders and target them with enforcement action (Himes and Crevier, 2021[58]). As it will be explored in Section 3 below, some jurisdictions have developed other alternatives such as ex-ante regulations with explicit bans.

Competition enforcement has also been recognised as another means to intervene against these practices. Because the use of misleading or deceptive online choice architecture practices may hinder switching22 and exploits consumer biases, they may result in a significant weakening of competition when consumers do not have the capacity to shift. In turn, they can distort businesses’ incentives, shifting competition away from product attributes that benefit consumers, such as quality and total price paid, towards features designed to induce purchasing decisions, including urgency cues or prominently displayed headline prices (CMA, 2022[54]; OECD, 2022[3]). Moreover, if enough businesses adopt similar practices, they could ultimately shape the market in a way that might further harm consumers, with competition among businesses being a “race to the bottom”23 (CMA, 2022[54]). When used by firms with market power, these practices can entrench this power and disadvantage new entrants, while reducing consumers’ trust in the market by preventing them from switching away. This, in turn, would raise rivals’ costs to persuade users to switch by competing for their attention through means other than competition (Himes and Crevier, 2021[58]). Thus, they have the potential to constitute abuses of a dominant position. For all these reasons, competition authorities have started looking at their effects on competition.

Scholars supporting antitrust intervention against misleading or deceptive online choice architecture practices argue that it is not a legitimate way of competing when digital companies build market power by manipulating consumers’ choices.24 They claim that to the extent that these practices impede consumers’ ability to select the best firms on the merits of their product offering, they can distort the competitive process as a whole. Moreover, they have identified that a path to consider misleading or deceptive online choice architecture practices as restrictive to competition is by understanding that these practices may significantly impede consumers’ autonomy as quality degradation25 (Day and Stemler, 2019[56]) with deceptive strategies generating exclusionary effects (Himes and Crevier, 2021[58]).

Behavioural considerations in competitive assessments have been a way to recognise and analyse misleading or deceptive online choice architecture practices in online markets as anti-competitive conduct and examine both exploitative and exclusionary effects. The European Commission’s decisions against Google (Google Shopping26 and Google Android27)also illustrate the relevance of behavioural evidence in this regard. The decisions highlighted consumer harm of certain online choice architecture practices (through algorithm biases) by drawing on behavioural evidence of the impact on consumers of ranking within search results and of the role of choice screens for search engine. In both cases, the Commission concluded that both rivals and customers were harmed.28

The intersection between the two policies in this context also lies in the fact that, in competitive markets, businesses that employ misleading practices would most likely be punished by consumers, by switching when they become aware of the practices. Hence increased competition could potentially reduce the risk of companies using deceptive practices, thus reducing the need for intervention. Recent studies have revealed, however, that consumers’ awareness in online markets is generally low and that behavioural biases mean that they underestimate manipulation in online more than offline contexts (OECD, 2022[3]). In order for the positive relationship to materialise without enforcement intervention, awareness initiatives towards consumers may therefore be required.

The analysis of personalisation and misleading or deceptive online choice architecture practices illustrates the nuanced interplay between consumer protection and competition objectives in digital markets. These practices do not merely generate isolated consumer detriment or efficiency concerns; rather, they reveal systemic dynamics whereby behavioural manipulation, informational asymmetries and market power can interact, amplifying potential harm. Moreover, the overlap between the two regimes underscores the importance of co-ordinated or complementary enforcement, where interventions in one domain can reinforce objectives in the other. This dual perspective not only clarifies the potential sources of consumer and competitive harm but also sets the stage for Section 3, which explores joint approaches and remedies to address complex digital‑market behaviour in a holistic and effective manner.

2.3.3. Other exclusionary practices

A number of business practices often reviewed by competition authorities under foreclosure theories of harm, due to their potential exclusionary effects, have also prompted scrutiny from a consumer protection perspective. This includes self-preferencing and tying and bundling, among other practices that have the potential to foreclose rivals.

Self-preferencing

Digital platforms rely on recommendations and rankings mechanisms to display results to consumers, for instance in response to product or service searches (Jürgensmeier and Skiera, 2025[59]). Self-preferencing practices, where digital businesses use their market power in one market to favour their own products in an ancillary market, are part of these strategies, and can have both pro-competitive effects and the potential to distort competition. As in any other situations involving a vertically integrated firm, self-preferencing may come as a consequence of objective product design decisions. It can be the result of competition on the merits, with the vertically integrated firm being the most efficient, thus facilitating consumers’ choice of products with better features (Jacobson and Wang, 2023[60]). Some literature analysing incentives of vertically integrated businesses, such as digital platforms, suggests that they may have stronger incentives to compete downstream, as they can partially internalise certain benefits of integration that are not available to independent firms (Katz, 2024[61]) but, overall, literature appears to be nuanced on both the positive and negative welfare implications of self-preferencing (Bowman and Prasad, 2025[62]).

Competition authorities have recently raised concerns that digital platforms with inherent conflicts of interest, acting both as intermediaries and as competitors to the sellers they host, may use self‑preferencing in ways that restrict competition when they hold a dominant position. These concerns are linked to intentionally raising rivals’ costs in order to gain significant competitive advantage in the market downstream leveraging from its dominant position as a platform.29 Although self-preferencing practices may also exist in offline markets,30 concerns in digital markets are greater particularly when these are prone to tipping, making it harder to restore competition once consumers have chosen their provider (Bowman and Prasad, 2025[62]). Investigations on behaviour by Amazon, Google and a number of other digital businesses have been opened in different jurisdictions in the past couple of years with sanctioning decisions consistently concluding that the self-preferencing conduct produced foreclosure effects.31 Moreover, ex-ante regulations have also been enacted in some jurisdictions, prohibiting self-preferencing in digital markets under certain conditions.32

Self-preferencing in a way that is not transparent to consumers (and especially when it is not linked to merits) may also be considered to violate consumer law by deceiving consumers. Where consumers reasonably perceive rankings or recommendations on an online marketplace as objective or relevance-based, but it has in fact prioritised its own products or services without disclosure, this may amount to a misleading commercial practice. Strategies such as fake reviews, manipulation of algorithms, or the opaque collection of data to adjust rankings or promote their own products, are part of self-preferencing practices that can and have been prosecuted by consumer protection authorities (Jacobson and Wang, 2023[60]). These practices are of particular concern in digital environments, where consumers depend on signals such as rankings and ratings to navigate large volumes of options and make informed choices. Linked to recommendations and rankings, opaque or confusing default or ordered results that can be perceived as objective recommendations, for instance as a result of undisclosed partnerships, can also result in consumer detriment, thus triggering action by consumer protection authorities (CMA, 2021[63]). Several jurisdictions have accordingly taken enforcement action33 or issued guidance34 requiring that the main parameters determining ranking be disclosed and that sponsored or paid placements be clearly distinguished from organic results.

Tying and bundling

Tying and bundling practices, which can be linked to self-preferencing as they can enable platforms to steer users towards their own services or integrated offerings, are also featured among competition intervention in digital markets. Tying practices involve businesses requiring consumers to purchase additional products or services alongside the product they wish to purchase and may arise either contractually (through obligations directed at consumers) or technically (through restrictions on interoperability) (OECD, 2024[6]; 2021[64]). In digital markets, the combination of high fixed costs and very low marginal costs makes tying a common strategy. By bundling products or services, platforms with relevant market power can strategically sustain negative prices, using tying as an implicit subsidy to recover fixed investments, strengthen their market power, compete more effectively and generate efficiencies from improved co-ordination across the different sides of the platform. Economies of scope can also generate pro-competitive effects of tying related to quality improvement from added “free” services (Wu and Philipsen, 2022[65]).

However, the anti-competitive effects of tying and bundling are particularly pronounced when implemented by firms holding a dominant position or substantial market power. These practices allow dominant firms to leverage their existing well-established position in certain markets to strengthen their presence in markets where they are less established, raising barriers to entry and potentially limiting competitive pressure. Thus, multiple recent enforcement investigations and decisions have focussed on tying practices and interoperability restrictions and their effects on competitors and consumers. Examples are investigations related to mobile operating systems (OS) and app stores where effects include harm to competition through significant competitive advantages gained using the tying practices that resulted in competitors not being able to compete effectively and with lower incentives to invest.35 Box 4 presents an example of parallel but complementary actions by the US Federal Trade Commission (FTC) and the Autorità Garante della Concorrenza e del Mercato (AGCM) against Amazon for practices related to its Amazon Prime service.36

Box 4. Enforcement actions in the United States and Italy against Amazon’s practices

In the United States, a jurisdiction where the US FTC has a dual mandate as both a consumer protection and competition authority, parallel and complementary actions have been pursued against different digital businesses for analogous conduct. One clear example are the actions against Amazon’s tying practices related to Amazon Prime. As discussed, while action in the framework of competition law explores exclusionary conduct, intervention in the framework of consumer protection focusses on effects on final consumers.1

In 2023, the US FTC together with 17 state attorneys sued Amazon for illegally maintaining its monopoly power by using a set of interlocking anticompetitive and unfair strategies. The complaint included allegations related to the use of anti-discounting measures and tying practices. With respect to the latter, the complaint states that Amazon’s practice of conditioning sellers’ ability to obtain Amazon Prime eligibility for their products on their use of its Fulfillment service² (for logistics) limited competitors’ ability to effectively compete against Amazon. The conduct, according to the complaints, also resulted in higher costs for sellers to offer their products on Amazon and other platforms.³ While most of the violations alleged referred to antitrust state and federal law, they also included violations of some States’ consumer law related to unfair practices and deceptive conduct. The case is ongoing.

Similar concerns were investigated in Italy. In 2024, the AGCM sanctioned Amazon for an abuse of dominance in which Amazon leveraged its position in the market for intermediation services on marketplaces to favour the adoption of Fulfillment, its logistic service, to the detriment of competing logistics operators. Among the practices introduced by Amazon, the company introduced benefits of Fulfillment users such as being eligible to use the Amazon Prime label.

1. Plaintiffs amended the complaint in October 2024, but the main arguments presented in this box remained; ². Amazon Fulfillment service is a logistics programme through which third-party sellers store their goods in Amazon’s warehouses, while Amazon manages storage, packaging, shipping, customer service and returns; ³. Plaintiffs amended the complaint in October 2024, but the main arguments presented in this box remained.

Sources: Complaint in Case No. 2:23-cv-01495-JHC, The United States District Court Western District of Washington; AGCM press release: A528 – Italian Competition Authority: Amazon fined over EUR 1.128 billion for abusing its dominant position, https://en.agcm.it/en/media/press-releases/2021/12/A528.

As tying or bundling involves some level of restriction on consumer choice, consumer detriment is clear, particularly when the practice is employed by a platform with significant market power and implemented deceptively. Examples could include situations where consumers are enrolled in bundled or add-on services without clear consent, where optional extras are presented as part of the main purchase without adequate disclosure, or where relevant interoperability limitations are not made clear.37 In these cases, tying practices also become object of consumer protection interventions. A recent illustration is provided by proceedings brought by the Australian Competition and Consumer Commission (ACCC) against Microsoft concerning its Microsoft 365 subscriptions. The authority alleges that, following the integration of its Copilot AI tool, Microsoft misled consumers by effectively presenting them with a choice between upgrading to a more expensive bundled plan or cancelling their subscription, while failing to clearly disclose the existence of a lower-priced alternative without the additional feature. The ACCC is seeking orders including penalties, injunctions, declarations, consumer redress and costs, as approximately 2.7 million consumers were allegedly affected by this conduct.38

The role of algorithms

The role of algorithms in facilitating exclusionary practices in digital markets also links competition and consumer concerns. The impact of algorithms on competition has been extensively discussed in previous OECD work (for a thorough review of the competition benefits and concerns of algorithm use, see (OECD, 2023[14]; 2025[50]).

Studies by the Competition and Market Authority (CMA) (2024[66]; 2021[63]), the Competition Bureau in Canada (Competition Bureau Canada, 2025[67]) and the Japan Fair Trade Commission (JFTC) (2021[68]), among others, have revealed that algorithms could facilitate the implementation of strategies aiming at limiting the entry or expansion of competitors, making these strategies more effective and less costly. This includes predatory pricing schemes, tying and bundling practices, self-preferencing practices and, more broadly, a variety of practices linked to the possibility to discriminate prices among consumers. Concerns are commonly linked to lack of transparency on these practices, mostly on how algorithms reach the prices or what the final price is; increased costs for certain customers, particularly the most vulnerable39 or competitors’ customers, and an overall increase in searching and transaction costs; loss of trust of consumers in the market; and potential exclusion of competitors linked to data misuse (CMA, 2021[63]).40

For example, in the case of predatory practices, pricing algorithms trained on detailed customers data can enhance the ability of a company to predate, by better identifying profitable opportunities in the predation and recoupment phases (OECD, 2025[8]). They can increase incentives to do so, by pricing in a way that minimises losses to the firm employing the strategy, reducing prices at a very specific level in which the strategy makes financial sense and targeting only certain customers, while accelerating recoupment once rivals exit (Competition Bureau Canada, 2024[69]). Similarly, price personalisation through the use of algorithms may make it easier for firms to leverage market power in one product to gain a competitive advantage in another by tying the two together, targeting discounts to identified consumers who place lower value on the tied product (Competition Bureau Canada, 2025[67]).

From a consumer perspective, algorithms may amplify informational asymmetries and behavioural biases, rendering offers less transparent, choices harder to compare and decisions more susceptible to manipulation. This has the potential of reducing consumer autonomy and undermining trust, in ways that closely mirror and reinforce the competitive risks (CMA, 2021[63]).

2.4. Conclusions

Overall, the practices discussed above illustrate how conduct in digital markets increasingly falls within the scope of both competition and consumer protection frameworks. The examples presented are not intended to provide an exhaustive list, but rather to highlight some prominent areas where the two policy domains intersect. In digital environments, certain practices may simultaneously affect consumer decision making and market dynamics, raising concerns that may be characterised as deceptive, exploitative or exclusionary depending on the analytical framework applied. While each policy addresses these practices through distinct legal standards and enforcement tools, their objectives often converge in protecting consumer welfare and ensuring well-functioning markets. This growing overlap suggests that a coherent understanding of digital market conduct may require closer co-ordination between the two policy domains. The following section explores how authorities have approached this interaction in practice and where further co-operation may be beneficial.

Notes

← 1.  The goals of competition policy were most recently discussed by competition authorities in 2022 in the framework of the OECD Global Forum on Competition.

← 2.  In 2020, the OECD Council adopted the Recommendation on Consumer Product Safety [OECD/LEGAL/0459] which recognises that businesses should only place safe products on the market, that consumers have a right to expect that products put on the market are safe under reasonably normal or foreseeable consumer use or misuse; and that compliance with product safety requirements by all economic operators can support a safe, fair and competitive consumer product marketplace.

← 3.  Such as efficiency, promoting innovation, or preserving opportunities for specific categories of business, like small and medium enterprises, in the case of competition law, or contributing to economic dynamism or empowering certain consumer groups, in the case of consumer law.

← 4.  While in some jurisdictions consumers are explicitly mentioned as a direct objective of the application of competition law, in others benefits to consumers are an indirect effect.

← 5.  At the same time, product safety obligations may themselves become a vector of competitive distortion when invoked strategically by dominant players to justify restrictions on rivals.

← 6.  For example, in 2025, during the process of implementation of a new consumer enforcement regime, the UK’s CMA explicitly recognised the need to streamline its guidance as “for smaller businesses especially, the compliance burden must be proportionate” (Cardell, 2025[96]).

← 7.  In Compass Banca SpA v Autorità Garante della Concorrenza e del Mercato (Case C-646/22, Judgment of 14 November 2024), the Court of Justice of the European Union clarified the interpretation of the “average consumer” standard under Directive 2005/29/EC on unfair commercial practices. While reaffirming the traditional notion of a reasonably well-informed, observant and circumspect consumer, the Court acknowledged that decision making may in practice be constrained by factors such as cognitive biases. Such factors do not automatically render a practice unfair, but national courts must assess whether, in the specific circumstances of a case, a commercial practice could materially distort consumer behaviour, reflecting an openness to integrating behavioural insights into the assessment of fairness and the validity of consumer choice.

← 8.  In addition to prices, businesses may use consumers data and profiles to disseminate personalised marketing output, such as advertising, offers and recommendations. With consumers data, companies are better able to predict how consumers will react to marketing and use the persuasion strategies that are most effective (Duivenvoorde, 2023[98]).

← 9.  For example, consumers with a willingness to pay that falls below the price that would be charged in the absence of personalisation.

← 10.  In some OECD jurisdictions, ex-ante legislations requiring transparency over personalisation offers that result from automated decision making have been introduced. Examples include Canada’s Directive on Automated Decision-Making (2019) and the EU Consumer Rights Directive (CRD) as amended by Directive (EU) 2019/2161 (The Omnibus Directive). More generally, data privacy laws include provisions broadly requiring transparency in data processing involving automated processes.

← 11.  Price discrimination has also been tackled by ex-ante regulation. While regulation is out of the scope of this paper, it is important to mention that across OECD jurisdictions, regulations have been issued aiming at ending unjustified discrimination in online markets. The clearest example is the EU ban on geo-blocking and the implementation of regulation (EU) 2018/302 prohibiting discrimination based on nationality, place of residence or place of establishment within the internal market. Other examples in the EU include the Commission's 2024 Digital Fairness Fitness Check and the forthcoming Digital Fairness Act.

← 12.  Although examples of unfair, discriminatory and excessive prices can be found in offline markets. For example, in Merci (Case C-179/90), the European Court of Justice determined that price increases granted by a dominant firm to certain consumers to offset price reductions to others can be unfair. Similarly, the CMA has conducted multiple price-cost analyses to review whether medicine prices have been anti-competitive (e.g. Case CE/9742-13).

← 13.  See, for example the discussion on proving capability of distorting competition in the CJEU ruling Case C-525/16 in the MEO case.

← 14.  In the United States, the FTC has used its powers to study markets and request information for research and policymaking purposes (under Section 6(b) of the FTC Act) to investigate personalised pricing, particularly with the use of personal data and algorithmic tools. In 2024, it issued orders to eight companies seeking information on price surveillance and has since issued summaries as intermediate outputs on its ongoing study. These summaries highlight considerations that are at the intersection of FTC’s responsibilities under competition, consumer and data privacy frameworks. See: https://www.ftc.gov/news-events/news/press-releases/2024/07/ftc-issues-orders-eight-companies-seeking-information-surveillance-pricing.

← 15.  One clear example is the Tinder case in Europe. In 2024, the European Commission and the Consumer Protection Co-operation (CPC) Network concluded an ongoing dialogue with the dating platform Tinder, which had implemented automated personalised discounts without explicitly informing users. To address concerns under EU consumer law, the platform committed to informing consumers clearly and upfront when pricing is personalised based on age and to disclosing that automated means are used to calculate discounts for premium services. Furthermore, Tinder agreed to provide the underlying rationale for these offers, such as identifying users who were previously unwilling to purchase services at a standard rate (European Commission, 2024[95]).

← 16.  See (Victor-Nyebuchi, 2025[97]; Canhoto, Keegan and Ryzhikh, 2023[92]; Mo et al., 2023[91]; CPRC, 2020[93]; Botta and Wiedemann, 2019[53]) for references on empirical evidence on consumers’ attitude towards personalisation and transparency about it. These studies show that what consumers in online markets dislike the most is the secrecy of personalisation rather than the personalisation itself, making it less clear that intervention against personalisation is needed (but transparency measures may be). This leads to preference for transparency and opt-out remedies rather than prohibitions. An additional consideration involves privacy concerns arising from consumers disliking being profiled (OECD, 2024[11]).

← 17.  OECD (2022[3]) presents a review of empirical evidence on the prevalence of misleading or deceptive online choice architecture practices, as well as on their influence on consumer decision making and their detectability.

← 18.  As classified by the (CMA, 2022[54]) in its taxonomy of online choice architecture practices. It is important to note, however, that only the manipulative character of these practices make them harmful for consumers as some online choice architectures can be beneficial.

← 19.  For instance, they can reduce transparency and disincentivise consumers to shop around, compare offers and switch. They can also influence consumers to purchase unsuitable products, choose inferior sellers and/or spend more than they want to. An OECD survey in 2024 revealed that online choice architecture practices significantly influenced online consumer decisions, causing significant financial, privacy and emotional impact. While the effects are on all consumers, older and infrequent internet users appeared more affected (OECD, Forthcoming[55]).

← 20.  Many OECD jurisdictions have provisions under their consumer laws prohibiting practices associated with deceptive, fraudulent or unfair online choice architecture. For example, Article 25 of the EU Digital Services Act (DSA) prohibits online platforms to “design, organise or operate their online interfaces in a way that deceives or manipulates the recipients of their service or in a way that otherwise materially distorts or impairs the ability of the recipients of their service to make free and informed decisions.” Similarly, Section 5 of the US FTC Act prohibits ‘‘unfair or deceptive acts or practices in or affecting commerce’’, applying to both online and offline markets.

← 21.  To cite only a couple of examples, (1) the US FTC secured a historic USD 2.5 billion settlement against Amazon, representing the largest ever civil penalty in an FTC case, for the use of deceptive methods to sign up consumers for Amazon Prime subscriptions while making it exceedingly difficult to cancel, Case No. 2:23-cv-0932-JHC; (2) there are current investigations opened against Temu in Europe, including by the European Commission for possible violations of the DSA, related to practices that would include fake discounts, pressure selling, forced gamification, fake reviews, hidden contact details and others https://ec.europa.eu/commission/presscorner/detail/en/ip_25_1913; and (3) the ACCC issued three relevant infringement decisions in 2024 to Dreamscape Networks International for practices linked to subscription traps and hidden fees for building company websites https://www.accc.gov.au/media-release/web-hosting-business-pays-penalties-for-allegedly-misleading-customers-about-%E2%80%98free-gifts%E2%80%99.

← 22.  For example, in its 2021 Digital Platform Services Inquiry, the ACCC identified different practices that hinder switching of browsers to Australian consumers.

← 23.  This refers to a situation in which businesses would progressively decline quality, transparency and/or other features of their products and services, which reduce consumer welfare.

← 24.  Deceptive practices as antitrust violations have been analysed in the past, including in offline markets (e.g. see (Harvard Law School, 2012[89])). However, the extent into which these practices could harm consumers in online markets have brought back these concerns to the debate.

← 25.  Quality degradation as a consequence of misleading practices was the main argument in Sherman v. Facebook, Inc. (3:20-cv-08721) where the plaintiffs argued that Facebook wrongfully acquired or maintained monopoly power by deploying “dark patterns, skullduggery, and other misleading and fraudulent behavior” in its data collection efforts causing consumers to give away more data and privacy than they would have otherwise, representing a lower quality of Facebook’s service.

← 26.  European Commission (EC) 2017, Case AT.39740 Google Search (Shopping).

← 27.  European Commission (EC) 2018, Case AT.40099 Google Android.

← 28.  More specifically, in Google Shopping, Google systematically positioned its comparison-shopping service more prominently in search results while demoting rivals. Given strong evidence that users disproportionately click on higher-ranked results, this self-preferencing diverted traffic from competitors, harming rivals (through loss of visibility and traffic) and consumers (through reduced choice and distorted comparison). In Google Android, Google imposed contractual restrictions tying Google Search and Chrome to the Android operating system and set them as defaults. Since behavioural evidence shows that consumers rarely change default settings, this foreclosed competing search engines, limiting rivals’ access to scale and ultimately restricting consumer choice.

← 29.  E.g. European Commission (EC) 2017, Case AT.39740 Google Search (Shopping) as discussed above.

← 30.  E.g. supermarkets promoting their own brands in-store as explained in (OECD, 2024[94]).

← 31.  See OECD (2024[6]) for a summary of enforcement action against self-preferencing in G7 jurisdictions and OECD (2025[90]) for self-preferencing investigations in Latin America that involved imposition or acceptance of remedies.

← 32.  For example, in Germany, self-preferencing practices are addressed in Art 19a (2) of the German Act against Competition Restraints (GWB). Similar provisions are contained in Article 5 of the EU Digital Markets Act (DMA).

← 33.  For example, in 2018 the CMA launched an enforcement action against hotel booking sites identifying concerns where rankings were influenced by commission paid to the site and later secured remedies to change those practices (see: https://www.gov.uk/cma-cases/online-hotel-booking). Similarly, in 2020, the Australian Federal Court found Trivago misled consumers where rankings gave significant weight to cost-per-click payments and often did not highlight the cheapest offers (see: https://www.accc.gov.au/media-release/trivago-misled-consumers-about-hotel-room-rates).

← 34.  In the EU, the Omnibus Directive explains that traders must provide information on the main parameters determining ranking. Moreover, the 2021 Guidance on the interpretation and application of Directive 2005/29/EC of the European Parliament and of the Council concerning unfair business-to-consumer commercial practices in the internal market states that providing search results without clearly disclosing paid advertisements or payments for higher ranking is prohibited.

← 35.  Ex-ante regulations enacted recently also tackles tying and bundling in digital markets directly, mostly in the shape of prohibitions. For example, in Japan, the Act on Promotion of Competition for Specified Smartphone Software (SSCPA) enacted in 2024 explicitly prohibits different tying practices such as tying a browser engine or payment services to the operating system.

← 36.  Amazon Prime is a paid subscription service that gives users access to additional services otherwise unavailable to other Amazon users. This includes benefits on delivery, streaming, shopping and reading services.

← 37.  The Google Android case illustrates an example of evaluating a firm’s capability to restrict competition through the lens of consumer behaviour. Notably, at paragraph 574 of the judgment, the General Court establishes the factual tendency of consumers not to download alternatives to preinstalled apps as the relevant context for this assessment. By dismissing Google's claim that users could easily remedy this by downloading a competing app, the Court affirms that tying practices exploit consumers’ "status quo bias". From a consumer protection standpoint, such conduct could be viewed as a constraint on decisional autonomy, as pre-set defaults can lead to consumers remaining with incumbent services even when alternatives are available. See: Case T-604/18 – Google and Alphabet v Commission (Google Android), https://infocuria.curia.europa.eu/tabs/affair?lang=en&sort=AFF_NUM-DESC&searchTerm=%22T-604%2F18%22&publishedId=T-604%2F18&juridiction=T

← 38.  See ACCC Press Release: Microsoft in court for allegedly misleading millions of Australians over Microsoft 365 subscriptions, October 2025, available at: https://www.accc.gov.au/media-release/microsoft-in-court-for-allegedly-misleading-millions-of-australians-over-microsoft-365-subscriptions.

← 39.  This includes personalisation generating and exploiting consumers’ susceptibilities such as insecurities, weaknesses and biases.

← 40.  In 2024, the OECD published a report on Algorithmic Pricing and Competition in G7 Jurisdictions, providing an overview of potential competition concerns commonly associated with their use (2025[50]).