INTRODUCTION
The move from traditional linear market structures towards multi-sided digital platforms has redefined the manner of operation of competition. Unlike traditional linear market structures, multi-sided digital platforms are positioned as intermediary platforms for different categories of users and create value based on cross-group network effects, where every additional user on one side of the market improves the overall experience for users on the other side of the market.1 This, coupled with the significance of data and the prevalence of zero-pricing mechanisms, which obscure the actual manner of monetizing consumer-side value, has led to an unprecedented level of market concentration among a handful of players, namely Alphabet Inc. (Google), Apple Inc., Meta Platforms Inc., and Amazon.com Inc., who are collectively referred to as digital “gatekeepers.”2 What sets these firms apart is not just their size, but also the nature of the infrastructure they control. Through their dominant positions in app distribution platforms, general search engines, social networking platforms, and e-commerce platforms, these firms are in a position to dictate the terms and conditions upon which dependent firms can access their end-users, a role which is not limited to any single relevant market.3 The real competition concern, therefore, is the ability of these firms to build upon and entrench their dominance in these adjacent markets, not just through self-preferencing and exclusion based on their proprietary data, but also through complex bundling dynamics, which is notoriously hard to address from an ex-post antitrust perspective. 4 Under the Indian legal framework, Section 4 of the Competition Act, 2002 is arguably the key measure for dealing with abuse of dominance. However, it is apparent that the section has not been fully developed to account for other forms of predatory practices, such as algorithmic exclusion, which are typical of Big Tech platforms’ market power conduct. In light of these developments, this paper seeks to provide a critical evaluation of Section 4 of the Competition Act and pose the following query: is the Competition Commission of India institutionally prepared for dealing with the structural abuses of digital gatekeepers? The identified gaps suggest a mismatch between the rate of digital foreclosure and the rate at which the extant framework is prepared to address it.5
CONCEPTUAL FRAMEWORK OF ALGORITHMIC PREDATION AND DIGITAL GATEKEEPING
A. Beyond Traditional Predatory Pricing
Traditionally, under Brooke Group Ltd. v. Brown & Williamson Tobacco Corp.,5 there are two necessary elements of a claim of predatory pricing: prices are set below some measure of costs, and recoupment of those losses through monopoly profits is reasonably likely. More specifically, under the Areeda-Turner formulation of costs,6 traditional predatory pricing assumes a competitive environment where competitive harm is price-proximate, where consumer welfare is quantitatively measurable, and where predation is limited in time. However, these assumptions have become theoretically unstable under the digital economy of algorithmic rules. Big Tech practices exclusionary conduct causing competitive harm without price reduction, practices it in a world of a zero priced economy where cost-based benchmarking is structurally inappropriate, and recoups through a multi-market data monetization approach rather than a post-predatory price increase in a discrete market. The concept of algorithmic predation is necessarily understood as a doctrinal category of exclusionary practices that are not necessarily or traditionally predatory and price-related, as it is practiced through algorithmic rules of platforms based on machine learning methodologies that systematically disadvantage competitors or create barriers to entry and enhancement of market position across a variety of interconnected relevant markets.7
B. Predatory Innovation and Algorithmic Collusion
There are two doctrinal developments in algorithmic predation that need to be conceptually clarified. The first one is the concept of predatory innovation, which is the utilization of product design changes, software, or platform changes by the dominant platform that are apparently pro-competitive but are implemented with strategic intent to reduce the overall quality, discoverability, and/or competitiveness of competing products.8 The apparent pro-competitive purpose of the product change in predatory innovation does not justify the change because the apparent pro-competitive purpose is not susceptible to explanation without the foreclosure effect. The other concept is algorithmic collusion, which is the capacity of individual pricing algorithms utilized by the platform to reach supra-competitive prices without the need to coordinate with other platforms, leading to the creation of cartel effects through the actions of apparently uncollusive algorithms that are not within the scope of the traditional requirements of collusion.9 Algorithmic collusion primarily engages with Section 3 of the Competition Act, 2002,10 and the relevance that algorithmic collusion holds for the assessment of the abuse of dominance doctrine is that it has the potential to pose questions for the mens rea assumptions that underlie the effects-based approach to the assessment of the abuse of dominance doctrine.
C. The Insufficiency of Price-Cost Tests in Digital Markets
The insufficiency of price-cost tests for digital platform conduct can be highlighted based on the following three structural features of digital markets. First, the zero price of the digital market for consumers categorically renders the price-cost tests insufficient due to the fact that
the absence of a price for the exchange renders the price-cost tests categorically insufficient, despite the fact that the platform is paid for the exchange by the consumer through the provision of data and attention.11 Secondly, the multi-sided nature of digital platforms implies that any pricing on one side of the platform effectively subsidizes the other side, and therefore, the application of cost allocation is not only conceptually meaningless but also analytically meaningless. Lastly, the mechanism of algorithmic exclusion is not dependent on pricing, and therefore, it is not within the scope of the Areeda-Turner test. Indeed, the OECD has already accepted the fact that the predation test, which is based on pricing, does not capture the full range of exclusionary practices that a dominant digital platform can indulge in.12 Further, the rationale followed in the decision-making process in the Google Android case by the Competition Commission of India effectively accepted the limitation of the price-cost test, as the findings in the case of abuse of dominance were not based on pricing considerations.23
D. Digital Gatekeepers in Multi-Sided Markets
The notion of digital gatekeepers is defined as a digital platform positioned in a structurally intermediary role between two or more distinct user groups whose interaction is facilitated, controlled, and monetised by the digital platform.13 In their seminal work on two-sided markets, Rochet and Tirole first identified and analysed the characteristics of digital platforms, where it was recognised that they internalise cross-group externalities and set prices on each of the two sides of the market to maximise total platform value.14 In digital platforms, this intermediary role gives the digital gatekeeper significant market power to condition access for advertisers, application developers, content providers, and merchants to user groups they would not be able to access efficiently through other means. This notion of digital gatekeepers is reflected in the European Union’s Digital Markets Act of 2022, where “gatekeepers” are defined based on quantitative criteria of size, user reach, and market position,15 a notion of market power that traditional antitrust law has had difficulty articulating and addressing beyond pricing.
E. Dual-Role Platforms: The Conflict of Vertical Integration
The competitive issue that is perhaps most representative of the type of gatekeeping engaged in by the dominant firms of Big Tech is the one associated with the operation of vertically integrated platforms that have the conflicting role of serving as the infrastructure for the access and competition of the platform’s competitors and, at the same time, engaging directly with these competitors in the downstream relevant markets. The operation of Amazon as both a platform for third-party sellers and a producer of private label products, as well as the operation of Google as both a general search engine and a provider of specialized search services, are examples of this type of conflict. The dual-role platform has the ability and the motivation to set the platform architecture in a way that benefits the downstream operations of the platform because the cost of this type of self-preferential platform operation is not reflected in the market prices that are subject to external scrutiny.
F. Self-Preferencing and Leveraging as Abuse
Self-preferencing, as used in the context of competition law, is the action of a dominant online platform, wherein the platform extends greater ranking, greater visibility, greater placement, and/or greater access conditions to its own downstream offerings than those of competing thirdparty offerings on the same platform.16 It is the harm to competition, rather than the harm to the quality of the self-preferenced product, which is the basis of the abuse. Leveraging, the overall abuse under which self-preferencing falls, is the extension of dominance in one relevant market into another adjacent market. Leveraging is, in fact, the only abuse under Section 4(2)(e) of the Competition Act, 2002.17 The ruling in the CCI’s decision on the Google Android case was a direct ruling on the issue of leveraging dominance in the licensable mobile operating system market into the markets of general web search services and web browsers, and thus represents the most analytically advanced application of the concept of leveraging dominance under Indian competition law.18
G. Network Effects: Direct and Indirect
Network effects are the principal structural driver of the self-sustaining dominance of digital platforms. Direct network effects are those where the value of the platform to a given user increases in proportion to the total number of users on the same platform side. This is exemplified in the dynamic of the communications network, as captured in Metcalfe’s Law.19 Indirect network effects, as seen in multi-sided platforms, are those where the growth of users on one platform side increases the value of the platform to users on the opposing side. This is exemplified in the dynamic where the growth of advertisers increases the value of the platform to consumers, who in turn increase the value of the platform to advertisers. The relevance of network effects in the assessment of dominance is found in their ability to create self-sustaining dominance, where a platform achieves critical mass on both platform sides, thus creating a structural dominance that new entrants cannot overcome through price and product differentiation, as the network effect represents an entry barrier of a fundamentally different kind from economies of scale.20
H. Data Feedback Loops and Competitive Entrenchment
The accumulation of data by dominant platforms also generates a compounding type of competitive advantage that operates independently of, and cumulatively with, the network effect. For instance, each user interaction with the platform generates behavioural data that improves the platform’s algorithmic performance, for example, the relevance of search results, the accuracy of recommendations, or the effectiveness of anti-fraud measures. Improved algorithmic performance, in turn, improves the user experience, which attracts more users and generates more data. This self-reinforcing type of competitive advantage, referred to as the ‘data feedback loop’ in the literature on competition, generates a type of dominance that is neither price-dependent nor structurally visible under traditional market share analysis.21 Perhaps more importantly, the data feedback loop generates entry barriers that are asymmetric because a new entrant cannot match the quality of outputs from an incumbent platform’s algorithms based on years of accumulated behavioural data, regardless of the capital investment and technical capability of the new entrant, because the data asset itself is non-replicable within the relevant timeframe of the competitive rivalry.
I. Zero-Pricing Models, Cross-Subsidization, and Market Tipping
The zero-pricing model, wherein dominant platforms charge zero monetary cost for services but collect user data and attention, selling these to advertisers, etc., performs two functions of competitive significance. First, the zero-pricing model locks out price-based competitors since new entrants cannot sustainably charge zero monetary cost without an equivalent data monetization strategy.22 Second, the zero-pricing model allows dominant platforms to crosssubsidize entry into new markets, using existing monopoly profits to undercut new entrants until they are forced out of the market, a strategy similar in economic effect to, but formally beyond, predatory pricing. Such factors ultimately give rise to the concept of market tipping, in which the process of competition dissolves to yield a ‘winner-take-all’ scenario as indirect network effects, data advantages, and switching costs collectively rise to the point at which multi-platform competition becomes economically viable to the user.23 Once a digital market has tipped, the conditions necessary to support meaningful competition cannot be addressed through behavioral responses, a reality that underpins the move to ex-ante forms of regulation that the Digital Competition Bill, 2024, now adopts.24
JUDICIAL LANDSCAPE UNDER SECTION 4 OF THE COMPETITION ACT, 2002 7.1. THE SCOPE AND STRUCTURE OF SECTION 4
Under Section 4 of the Competition Act, 2002, “no enterprise or group of enterprises shall abuse its dominant position in the relevant market.”25 The provision appears to operate through a two-stage analytical framework, wherein the authority must first establish dominance on the part of the enterprise in the “relevant market,” and then establish whether the “abuse” on the part of the enterprise falls within the parameters of the “abuse” envisaged under Section 4(2). Dominance, as defined under the Explanation appended to Section 4, refers to “a position of strength enjoyed by an enterprise in the relevant market which enables it to operate independently of the forces of competition, or to operate in its favor,37 in relation to its competitors, consumers, or the relevant market.” It is pertinent to note, however, that dominance does not require the existence of monopoly or percentage shares, but rather appears to be based on the functional concept of dominance. The criteria for determining dominance is set out under Section 19(4) of the Act, where a non-exhaustive list of criteria is set out, which includes market share, size and resources of the enterprise, size and significance of competitors, economic power and any commercial advantage over competitors, vertical integration, dependence of consumers, countervailing buying power, and market structure and size of the market.26 In its digital market jurisprudence, the Competition Commission of India has selectively dealt with these criteria, at times prioritizing market share analysis over structural entry barrier analysis, a methodology that has led to dominance analyses of varying analytical depth and rigor, especially in platform market settings where market share figures are contentious and dynamic.
The five types of abusive behavior are enumerated in Section 4(2). Section 4(2)(a) prohibits the imposition of unfair or discriminatory terms or prices in the purchase or sale of goods or services. Section 4(2)(b) prohibits the restriction or limitation of production, markets, or technical development to the prejudice of consumers. Section 4(2)(c) prohibits conduct that leads to the denial of market access. Section 4(2)(d) prohibits the conditioning of contracts on the performance of other obligations that have no relation to the subject matter of the contract, the tying provision. Section 4(2)(e), most relevant to digital market abuse, prohibits the use of dominance in one relevant market to enter another relevant market, the leveraging provision.27 The CCI has applied these provisions with variable degrees of analytical rigor in its digital cases, and a close reading of its decisions reveals that there is a tension between the textual scope of the provisions and the structural nature of platform-based exclusionary conduct that does not easily fit within the boundaries of one or another of the provisions.
The CCI has clearly adopted an “effects-based approach” in assessing abuse, wherein it assesses the competitive harm caused by the impugned conduct rather than adopting a per se approach of determining which conduct is or is not abusive.28 This approach, theoretically consistent with modern economics of competition, requires the CCI to establish a “credible theory of competitive harm,” “identify a relevant counterfact,” and “evaluate whether competitive foreclosure is of a magnitude that cannot be justified by pro-competitive efficiencies.” However, in digital market cases, the application of such an “effects-based approach” is subject to a structural evidentiary constraint in that the anticompetitive mechanism is often contained in “proprietary algorithms whose parameters are not available to the CCI without compelled disclosure.” The verification of competitive harm is therefore “analytically contingent on respondent-supplied evidence.” This issue of opacity of algorithms has not been examined in CCI jurisprudence, and it is yet to be seen whether a coherent evidentiary framework is developed for algorithmic conduct without reliance on inference or respondent concession.
A. Landmark Indian Digital Market Cases
1. Google Android Case (2022)
The CCI’s decision in the case of “In Re: Allegations of Anti-Competitive Practices by Google LLC”29 can be regarded as the most analytically interesting application of Section 4 to digital platform activities from the Indian competition law perspective. The case involved the business practice followed by Google of conditioning the licensing of the Android mobile operating system to device manufacturers on the imposition of a series of anti-competitive obligations, which included the mandatory pre-installation of Google Search and Google Chrome, antifragmentation obligations that prohibited the manufacture of Android forks, and revenuesharing arrangements contingent upon the provision of exclusivity for the placement of Google Search. The Commission identified three different relevant markets: the market for licensable mobile operating systems, the market for app stores for Android mobile operating systems, and the market for general web search services in India.30 With its dominance established in all three markets, the CCI found that the pre-installment obligation and anti-fragmentation constituted tying under Section 4(2)(d) and leveraging under Section 4(2)(e) of the Act as it leveraged its dominance in the licensable mobile operating systems market to maintain and build upon its dominance in the web search and browsers markets.31 The CCI also found that the revenue-sharing agreements between Google and the device manufacturers on the condition of Google Search exclusivity constituted a denial of market access to competing search engines under Section 4(2)(c). The doctrinal importance of the decision lies in the CCI’s recognition that the tying effects in digital platform markets are driven by the ecosystem architecture, not the contractual coercion that the traditional notion of tying might suggest, and that the relevant harm is not the price distortion that is visible to the consumer but the foreclosure effects that allow competitors to achieve the requisite distribution base to compete. The CCI imposed a fine of approximately ₹1,337.76 crore and ordered Google to allow device manufacturers to pre-install third-party applications without the need to bundle the same.32 The decision has been subject to various appellate proceedings, including the National Company Law Appellate Tribunal, which declined to stay the remedial directions, and the Supreme Court, which has finally upheld the relevant directions.¹⁰ The journey that the case has taken to reach the Supreme Court points to the inherent problem with the Indian framework, which is that the time lag between the infringement and the remedial directions is so large that, in the context of digital markets, the relevant dominance could be consolidated before the directions are implemented.
2. Meta/WhatsApp Privacy Case
The proceedings of the CCI in the WhatsApp policy update case have thus added a new doctrinal layer that the Indian competition regime has not been called upon to address at a deeper level in its jurisprudence till now that of the competitive significance of data collection as a criterion for non-price competition.33 The investigation conducted by the Commission suo motu in response to WhatsApp’s mandatory update to its user policy that required users to agree to share expansive amounts of data with Meta Platforms Inc.’s corporate family as a condition for continued access to its platform raises the question of whether the imposition of expansive data collection obligations by a dominant messaging platform constitutes an unfair condition in the context of Section 4(2)(a) of the Competition Act, 2002.34 The analytical framework that is needed for addressing this question calls for a reconceptualisation of consumer harm beyond price and output considerations. When the platform service is zero-priced, the competitive harm is reflected in a worsening of privacy as a quality dimension of service, which is a form of non-price exploitation that is structurally beyond the capacity of price-referential abuse tests to evaluate. The CCI’s determination that WhatsApp’s imposition of a demand for data sharing on its users constituted an unfair condition on users of its service who did not have a practicable alternative platform choice in light of the network effect-based lock-in of users on its platform marks a significant development in that it confirms that data exploitation by dominant platforms is within the abuse of dominance framework even in the absence of monetary price harm.47 The case also confirms that privacy and competition law intersect in digital platforms at the point at which data collection is simultaneously consumer exploitation and a means of creating and sustaining data-based dominance against new market entrants.
3. Apple App Store Investigation
The CCI’s examination of Apple Inc.’s business practices in relation to the Apple App Store, based on information filed by various app developers, reveals structural concerns similar to those analyzed in European proceedings under Article 102 TFEU and the European Union’s Digital Markets Act.35 Apple Inc.’s business practices under examination include the 30% mandatory commission charged on in-app purchases facilitated through Apple’s proprietary payment system, anti-steering provisions, and the walled-garden ecosystem architecture, which limits iOS application distribution exclusively to the Apple App Store. In analyzing the competition law implications, there is a need to consider the concept of the aftermarket, where Apple Inc.’s dominance in the iOS app distribution market, which is dependent on Apple’s hardware business, allows Apple Inc. to charge supra-competitive commissions on the Apple App Store, where app developers have no alternative but to transact on the Apple platform.36 The anti-steering clauses add an additional layer of foreclosure, as they prevent developers from informing users of cheaper alternatives available through other sources, and shield Apple’s payment processing monopoly from competitive pressure. Although it is very early days for the CCI investigation, it is likely that the more detailed structural analysis required, especially in relation to defining the App Store as a bottleneck facility and the commission as an unfair method under Section 4(2)(a), will require a more direct engagement with these issues than its existing case law has asked of it.
B. The Price-Cost Dilemma in Digital Markets
The basis for the adjudication of cases of predatory pricing under the conventional principles of competition law is the analytical model propounded by Areeda and Turner, which holds that pricing at less than the average variable cost constitutes a prima facie case of predatory pricing because it would not be rational for a profit-maximizing firm to engage in such pricing unless it were for the purpose of predation and subsequent recoupment of the losses at monopoly prices.37 The Brooke Group test, as propounded by the United States Supreme Court, adds the element of recoupment to the pricing at less than cost test and holds that claims for predatory pricing will not be cognisable unless there is a dangerous probability of recoupment of the investment made by the dominant firm in pricing at less than cost through the exercise of monopoly power following the foreclosure of the competitor.38 Under the Indian competition law, the Explanation to Section 4 of the Act defines predatory pricing as the sale of goods or the provision of services at a price lower than the cost as may be determined by the regulations with the object of reducing competition or eliminating a competitor.39 The application of average variable cost tests in assessing digital platform conduct is subject to a series of structural objections that collectively negate the doctrinal validity of predation tests based on reference to price in zero-price platform markets. First, the fundamental assumption of the Areeda Turner framework that below-cost pricing is indicative of predation because it is economically irrational and therefore inconsistent with short-run profit maximisation is simply not applicable in multi-sided platform markets in which a platform rationally sets zero or negative prices on one side of its platform in order to maximise total platform value through cross-platform network effects.40 The zero-price offered by Google on its search services, by Meta on its social networking platform, and by Amazon on its Prime video platform does not evidence predation but simply illustrates the economic reality of platform pricing in two-sided platform markets in which revenue on the consumer side of the platform is sacrificed in return for the monetisation of advertising inventory on the other side of the platform. The mechanical application of the below-cost pricing test to such platforms would yield a positive predation finding in circumstances in which such pricing is economically rational and pro-competitive and pro-consumer.
Secondly, the cross-subsidization of different sides of multi-sided platforms makes it analytically impossible to calculate average variable costs across the multi-sided platforms as a whole without a principled methodology for allocating joint costs across user groups who receive fundamentally different and non-comparable forms of platform value. Furthermore, the CCI’s regulations on cost determination41 fail to address multi-sided platforms’ cost allocation, and there is thus no coherent analytical instrument available for application of the statutory standard of predatory pricing by the Commission.42
Lastly, and most fundamentally, algorithmic exclusion tends to happen through mechanisms entirely independent of pricing conduct, and search ranking manipulation, app store demotion, interoperability degradation, and access restriction of digital platforms inflict competitive harm on rivals without any price reduction for consumers. Thus, the very framework of predatory pricing is entirely inapplicable to algorithmic exclusion, as it is not based on any price reduction mechanism. This is a limitation of price-cost tests in algorithm-driven digital ecosystems, and it is a structural limitation of Section 4 of the Act, one that is beyond remedying through any doctrinal approach and is one against whose adequacy any legislative reform of the Act through the Digital Competition Bill, 2024, is necessarily assessed.
THE REGULATORY SHIFT – FROM EX-POST TO EX-ANTE
THE DIGITAL COMPETITION BILL, 2024
The rationale for the Digital Competition Bill, 2024 is rooted in the structural understanding, formally articulated in the Committee on Digital Competition Law Report of 2023, that the expost enforcement framework of the Competition Act, 2002 is institutionally not suited to deal with the pace, opacity, and self-reinforcing nature of exclusionary conduct in digital markets.43 The CDCL Report, constituted under the aegis of the Ministry of Corporate Affairs, formally recognized that “the temporal gap between exclusionary conduct and its remedial consequences is structurally incompatible with digital markets because dominance and tipping occur on timelines that ex-post enforcement cannot remediate.”44 The CDCL Report drew upon the experience of the European Union’s Digital Markets Act, the Digital Markets, Competition and Consumers Act of the United Kingdom, and the recommendations of the Competition Law Review Committee to formally recommend that “a separate ex-ante regime, alongside and independent of Section 4, is needed to ensure contestability in the digital economy of India.”45
The underlying regulatory logic of the Bill, as it sets out to establish new rules, marks a change from the effects-based approach of Section 4 in one respect that is philosophically significant. While the effects-based approach under Section 4 requires the investigating authority to establish harm, or likely harm, that is attributable to the competitive conduct of a dominant undertaking, the proposed approach under the Digital Competition Bill proposes to impose prospective obligations on designated undertakings as a matter of structural necessity, quite irrespective of harm. This change from harm-based enforcement to status-based obligation reflects, in my view, a judgment on the part of the legislature that the very structural position of gatekeepers in the digital economy is, in itself, a matter of regulatory concern quite apart from the very exercise of those powers, although, as noted, this gives rise to important constitutional and proportionality issues that must be addressed.
The Bill establishes the notion of Systemically Significant Digital Enterprise as a key notion on which ex-ante obligations are applicable.46 Being considered a SSDE is proposed to be based on a set of quantitative criteria, such as turnover in India, global turnover, market capitalisation, and end users and business users in India, as well as a qualitative evaluation of its significance as a gateway or bottleneck for business users looking to reach end users.47 This two-pronged approach is similar to what is proposed under the EU DMA, where gatekeepers are identified based on a combination of quantitative presumptions and qualitative evaluation.48
The mechanism of designation also raises a key threshold concern of principle: that the Bill as drafted does not clearly articulate the standard of proof and analytical approach by which the designated authority will evaluate “systemic significance” beyond numerical thresholds, leaving the risk of arbitrary and over-inclusive designation of platforms whose structural characteristics fail to confer the gatekeeping potential on which such regulation is intended to focus. The lack of a formal contestation mechanism of adequate procedural sufficiency in designation, akin to the rebuttable approach in Article 3(5) of the DMA, also raises concerns regarding natural justice and adequate administrative process concerns with the Bill.62
The specific obligations for SSDEs under the Bill include the prohibition of self-preferencing, anti-steering, tying, and bundling, as well as the use of non-public business user data for competitive advantage and the limitation of interoperability and data portability.49 The obligations for SSDEs, from a structural viewpoint, amount to a codification of the various forms of abuse that have been difficult to address under Section 4, and their translation into a system of compliance based on rules rather than harm. The doctrinal importance of this transfer from one system to the other should not be underestimated: an enterprise under SSDE obligations cannot plead the efficiency argument or the pro-competitive effect as it would be evaluated under the abuse assessment system.
This categorization, while creating greater certainty and decreasing latency in enforcement, produces a tension with the innovation incentive-based arguments that dominant digital platforms have consistently offered throughout competition proceedings. The imposition of obligations of interoperability and data sharing on identified platforms risks, if not economically precisely calibrated, reducing the private return on investment for platform infrastructure and thus undermining the dynamic efficiency-based justification for permitting platforms to scale in the first instance.50 The CDCL Report recognizes this tension but does not address it with any level of doctrinal specificity, noting broadly that obligations should be “calibrated and proportionate” without any framework for what is meant by proportionality for purposes of contested designation proceedings.51
The constitutional viability of the Digital Competition Bill’s identification and obligation system raises implications for Articles 14 and 19(1)(g) of the Constitution of India. The imposition of categorical obligations on a class of undertakings based on structural significance and size without individualized harm analysis will have to meet the reasonable classification test as prescribed by Article 14 of the Constitution.52 The reasonable classification test demands a rational nexus between the classifying feature and the legislative purpose. Although the rational nexus between structural gatekeeping powers and potential competitive harm is analytically established in economic theory on digital competition, the absence of proportionality analysis in the administrative system of the Digital Competition Bill to assess whether obligations imposed on undertakings are no more burdensome than necessary to meet the regulatory purpose creates a vulnerability that will be exploited in constitutional adjudication. The proportionality analysis of the Digital Market Regulation of the EU by the EU’s highest court will be relevant to assess the constitutional viability of the Digital Competition Bill’s identification and obligation system. Although not binding on the courts in
India, it can be used as an analytical tool to assess the constitutional viability of the Digital Competition Bill.53
B. Comparative Analysis
1. European Union – Digital Markets Act, 2022
The EU’s Digital Markets Act, 2022, is the most developed doctrinally ex ante digital market regulation in force, and the structural framework is the most directly comparable framework for assessing the Digital Competition Bill.54 The EU’s DMA categorizes “gatekeepers” using quantitative criteria such as EEA turnover of at least 7.5 billion euros, a market capitalization of at least 75 billion euros, at least 45 million monthly active end-users in the EEA, and at least 10,000 yearly active business users, along with qualitative criteria such as entrenchment and durability.55 The Commission has designated Alphabet, Apple, Meta, Amazon, Microsoft, and ByteDance as gatekeepers, and this triggers the full range of obligations under Articles 5 and 6 of the Regulation.56 Article 5 sets forth “black list” obligations, which are non-justiciable and include restrictions against personal data accumulation across core platform services without consent, against restricting business users from directing consumers to their own platforms, and against requiring messaging services to become interoperable with third-party providers.57
Article 6 sets forth “grey list” obligations subject to regulatory dialogue and specification, including obligations to prohibit self-preferencing in product ranking, non-public business user data in competition with business users, and the need to provide effective data portability.58 The structural distinction between the per se prohibitions in Article 5 and the specifiable obligations in Article 6 reflects a nuanced regulatory design that the Digital Competition Bill has not yet achieved with equivalent clarity and that should be directly adopted in the legislative design of the Digital Competition Bill.
The DMA’s enforcement structure confers “exclusive competence” to the European Commission for non-compliance with a penalty of up to ten percent of global annual turnover for non-compliance and up to twenty percent for systematic non-compliance, with the added threat of structural remedies for systematic non-compliance.59 The enforcement experience of the DMA in its first phase of operation, including proceedings against Apple browser choice screen obligations and Meta’s consent or pay model, highlights the regulatory ambitions as well as the institutional costs of ex-ante digital market enforcement that the CCI may be structurally incapable of meeting without equivalent augmentation.60
2. United States – American Innovation and Choice Online Act
The American Innovation and Choice Online Act, proposed in the United States Congress, aimed to bar the platforms from self-preferencing, discriminatory treatment of competing products, and restricting interoperability for platforms with at least 50 million active monthly users or 100,000 active monthly business users and a market capitalization or net annual sales of $550 billion.61 The anti-self-preferencing provisions of the Act were conceptually similar to the DMA’s Article 6(5) obligations. The provisions marked a notable change in the United States’ longstanding reliance on the Sherman Act’s monopolization standard, which demands evidence of wilful acquisition or retention of monopoly power through exclusionary practices subject to a stringent effects-based analysis.62
The AICOA was not enacted, failing due to resistance in the legislature from supporters of the platform industry and concerns about the scope of the bill and due process concerns regarding the categorical prohibitions in the bill. The failure of the Act in the United States is, from a theoretical perspective, instructive for the proponents of the Digital Competition Bill in India, as it shows that the political economy of digital regulation is just as important as the substantive rules and that any framework based upon categorical obligations for digital platforms requires a level of legislative agreement about the balance between structural competition concerns and innovation incentives, which is not present in the American political context, in contrast to the European framework. The Indian context, in which competition regulation is undertaken by parliamentary enactment and in which the federated legislature is not present, is more conducive to ex-ante regulation, although the proportionality concerns raised in the context of the AICOA are, from a theoretical perspective, valid regardless of the context.63
C. The “Black Box” Challenge
The most structurally intractable enforcement challenge facing both the existing Section 4 framework and the newly proposed Digital Competition Bill framework is arguably the fundamental opacity of the proprietary algorithms employed by dominant digital platforms in their management of search engine ranking, content curation, advertising allocation, and product recommendations. The concept of algorithmic opacity can be defined as the state in which the underlying computational logic determining the outputs of a digital platform, and thus the mechanism by which self-preference or exclusion is pursued, remains inaccessible to verification and evaluation outside of forced disclosure due to the proprietary nature and complex parameters of the algorithmic design, as well as the inherently complex and technical nature of such algorithms, and their self-modifying nature in response to machine learning processes.64
The evidentiary burden that such opacity places on competition enforcement authorities is significant. In order to demonstrate self-preferencing of algorithms pursuant to Section 4(2)(e) and its equivalent provisions of the Digital Competition Bill, the CCI must not only show that the results of a dominant platform’s search engine or its outputs on product ranking favor its own products, but also that such favoring is caused by its algorithms and not because of superior inherent relevance and quality of its own products. This counterfactual causality is a complex inquiry that the CCI’s existing investigative capabilities are not yet suited to conduct, as they are based on document discovery and examination of witnesses and economic analysis of market outcomes.65
The Competition Act of 2002 grants power to the Director General for investigation, information collection, and inspection of premises but fails to specifically address the issue of technical auditing of algorithmic systems and the appointment of independent technical experts with access to proprietary code of platforms.66 This deficiency of the investigation framework suggests that the CCI’s penetration of algorithmic complexity is subject to voluntary cooperation of respondent platforms or inferential power of observed market outcome-based data, where respondent platforms have exploited this limitation of the methodology of investigation in proceedings by claiming algorithmic design decisions as pro-competitive optimising decisions beyond independent evaluation of respondent platforms.
The framework of the Digital Competition Bill is not sufficient for addressing the issue of investigation of algorithmic complexity of platforms. For ex-ante monitoring of compliance of SSDE obligations regarding self-preferencing and data use restrictions, it is necessary for the designated authority to constantly assess and verify whether algorithmic conduct of platforms is aligned with mandated obligations. However, it is technically impossible for them to do so.67
The right of the European Commission to conduct technical access inspections and appoint independent technical experts under Article 26 of the DMA,68 with its provisions for independent auditing obligations of designated gatekeepers, offers a model of regulatory design that the Digital Competition Bill is obliged to follow if its ex-ante obligations are to mean anything more than mere window-dressing.
The challenge to the design of regulation in this area is ultimately as much an institutional as a doctrinal one. The CCI’s current technical capabilities do not extend to a standing body of algorithmic engineers, data scientists, or platform technology experts that can engage in independent algorithmic audits or assess the competitive implications of machine learning model design choices.69 Without a substantial investment in technical expertise of the sort that the European Commission’s Digital Markets Unit has undertaken in preparation for the DMA’s impending enforcement landscape, the ex ante obligations framework set out in the Digital Competition Bill risks falling into a compliance theatre model in which platforms merely go through the motions of compliance without fundamentally changing the algorithmic behavior that the obligations were intended to correct.
The countervailing risk of over-regulation must be appreciated with the same degree of analytical seriousness. Algorithmic audit requirements, if not drafted with sufficient specification with regard to scope, methodology, and confidentiality, may necessitate the disclosure of truly proprietary technical information that is competitively sensitive but unrelated to the platform behaviour under regulatory assessment, thereby creating a chilling effect on platform innovation investment that is socially costly relative to the regulatory benefits obtained. The appropriate response to this problem, both practically and institutionally, is not to eschew algorithmic audit but to design the framework with procedural specificity, such that compelled disclosure is restricted to the algorithmic parameters relevant to the compliance with the designated obligation, subject to review by independent technical experts with regard to confidentiality, which is a design specification that the Digital Competition Bill must achieve if the legislative framework is to successfully navigate the regulatory spectrum between underdeterrence and over-deterrence with conceptual clarity.70
FINDINGS AND GAPS
A. Doctrinal Inadequacy of the Section 4 Framework in Digital Markets
The analytical exercise undertaken throughout the preceding chapters of this work establishes a cohesive and cumulatively supportive doctrinal argument, namely, that the abuse of dominance framework under Section 4 of the Competition Act, 2002, despite its structural capability of incorporating digital practices under its textual categories, is nonetheless analytically inadequate for addressing the systemic and structurally self-reinforcing nature of Big Tech gatekeeping practices. Such analytical inadequacy is not merely evidentiary and institutional but rather foundational and intrinsic, grounded as it is in a framework originally designed for linear and price-competitive single-homed markets and applied unquestioningly and unmodified to multi-sided digital ecosystems whose very market dynamics operate according to entirely different mechanisms.
The price-cost test framework, as informed by the Areeda-Turner approach and partially reflected in Indian competition law through its partial incorporation under the predatory pricing definition of Section 4 of the Act under the Explanation thereto, is categorically inadequate for digital markets where price is zero and where consumer-side monetization is deferred, crosssubsidized, and/or absent in its monetary expression.71 The recoupment requirement of the Brooke Group standard, itself a doctrinal device intended to screen out economically irrational predation claims, becomes analytically impossible when applied to platforms recouping investment not through post-predation price increases in one market, but through data collection, advertising monetization, and lock-in across multiple interconnected relevant markets simultaneously. No doctrinal adjustment, it is submitted, less than a fundamental rethinking of what constitutes competitive harm in zero-price multi-sided markets, can cure this analytical impossibility under the existing legislative framework.
B. Algorithmic Self-Preferencing and the Evidentiary Gap
The CCI’s consideration of algorithmic self-preferencing under Section 4 of the Act in its Google Android proceedings is arguably the most advanced application of Section 4 doctrine in relation to platform leveraging under Indian competition law, yet at the same time, its reasoning arguably highlighted the evidentiary limitations of algorithmic practices for assessing effects-based abuse under Section 4(2).72 Essentially, the direct relationship between algorithmic design and market foreclosure, which had to be established for the CCI to sustain its findings of leveraging under Section 4(2)(e), was arguably based on market outcome-based evidence and economic reasoning rather than any direct analysis of Google’s search engine ranking algorithms. Although arguably sustainable from a legal perspective, it is arguably a very fragile evidentiary framework for any future algorithmic self-preferencing cases, particularly where more sophisticated forms of self-preferencing might not lend themselves as easily to outcome-based reasoning and inference.
C. The Structural Case for Ex-Ante Regulation
The cumulative doctrinal findings that have been arrived at in this paper are supportive of the conclusion that the effectiveness of ex-post enforcement under Section 4, optimally implemented, is structurally not capable of re-establishing competitive market conditions in digital markets that have already tipped towards dominance. Network effects, data feedback loops, and switching cost asymmetries create self-reinforcing market dominance that the behavioural remedies under Section 27 of the Competition Act are not capable of countering within commercially meaningful time periods.73 The Digital Competition Bill, 2024, in its proposal to create a regime of ex-ante obligations for Systemically Significant Digital Enterprises based on the principle of designation, is a legislative recognition of the limitations of the effectiveness of ex-post remedies that is analytically well-supported, though the proportionality, procedural, and institutional issues that the Bill raises are many, the Draft Bill notwithstanding. The experience with the EU’s Digital Markets Act, which has served as the basis for the Digital Competition Bill, 2024, is supportive of the regulatory, procedural, and institutional challenges that the regime of ex-ante digital market enforcement poses, which the Indian framework must address with specificity to achieve the contestability goals that the proposed regime is intended to achieve.
PROPOSED REFORMS
A. Implementing Algorithmic Audits Within the Competition Act Framework
The most immediate need in terms of amending the competition enforcement framework in India is the incorporation of a formal algorithmic audit mechanism, whereby the Director General is empowered to commission technical audits of proprietary algorithms in cases of alleged self-preferencing, ranking, and data-driven exclusion. The legal foundation for such a mechanism already exists in the Competition Act in a nascent form, whereby Section 41 grants the Director General the ability to seek any information or document deemed relevant to the case, and Section 36(2) grants the Commission the ability to engage technical experts for the purposes of proceedings.74 Clear legislative provisions regarding algorithmic audit authority, including access to algorithmic design parameters and documentation of ranking methodology, would establish this investigative tool on a firm footing and clarify the current uncertainty regarding the extent of technical access powers vested in the Director General. The issues of proportionality and due process in relation to algorithmic audits are legitimate concerns and will need careful addressing in the design of the legislation in detail. Compulsory disclosure of algorithmic architecture is a form of interference with proprietary intellectual property rights, and it is important that provisions are made for confidentiality and disclosure of information only on parameters relevant to competition issues after review by technical expert review under a closed procedure and access restricted to relevant parameters by the Commission in its investigation role and not as a tool for gleaning proprietary technical information unrelated to competition issues. A requirement of proportionality review, where audit directions are issued only where the Commission establishes a prima facie case of algorithmic abuse and identifies the specific algorithmic parameters relevant to the alleged abuse, would help to better align the mechanism with constitutional due process principles under Article 21 and principles of natural justice.75
B. Redefining Relevant Market in Multi-Sided Platform Contexts
The methodology of market definition adopted by the CCI for digital platforms involves a need for doctrinal reformulation of the approach, taking into account the inherent characteristics of digital platforms, which are misrepresented under the SSNIP test and substitutability analysis. However, the proposed reform is not one of abandoning the notion of a relevant market as an analytical framework, as it would create an element of indeterminacy in dominance analyses that is not compatible with the rule of law demands of competition law adjudication, but rather one of reconceptualising it along ecosystem lines.76
An ecosystem-based market definition will involve assessing dominance not only within a narrowly defined product or service market on one side of a platform but also across the entire ecosystem of markets in which the structural position of the platform with respect to user access and interoperability generates competitive effects. This will involve the CCI having to factor in cross-side network effect analysis in its Section 19(4) dominance test and treating the platform’s control of complementary user groups as a structural barrier to entry of equivalent competitive concern to traditional economies of scale. The regulations framed by the CCI under Section 64 of the Competition Act could be amended to set out guidelines on ecosystem-based market definition methodology without requiring primary legislation and therefore be a relatively expeditious means of achieving doctrinal change within the existing framework.77 Moreover, it is essential that data centrality is specifically built into the framework of dominance assessment as a separate element of market power, alongside other Section 19(4) criteria. The recommendation of the CDCL Report regarding the acknowledgment of benefits related to data as an element of dominance in digital markets is a necessary doctrine that is to be implemented through formal regulatory guidelines outlining the methodology for analyzing entry barriers related to data, data accumulation as a structural advantage, and the competitive significance of non-replicable proprietary data assets in digital markets.92
C. Institutional Strengthening of the Competition Commission of India
The above-referenced reforms to the relevant doctrine can be implemented only to the extent that the CCI has the institutional capacity to do so with technical rigor and analytical credibility. The present institutional structure of the Commission, comprised in the main of legal and economic professionals without technical specialisation in platform engineering, algorithm systems, or data science, is not adequate to meet the demands of enforcing digital market competition regulation as analytically identified in this paper.78 The creation of a Digital Markets Unit in the Director General’s office with technical specialists in algorithm systems, platform architecture, and data analytics represents a fundamental change in the structure of the Commission that cannot be delayed without undermining the effectiveness of both the existing Section 4 regime and the proposed Digital Competition Bill.
Coordination between the CCI and sectoral regulators such as the Data Protection Board, established under the Digital Personal Data Protection Act, 2023, and the Telecom Regulatory Authority of India, is another key institutional imperative, with the subject matter of digital market competition frequently involving areas of regulatory overlap between data protection, consumer protection, and telecommunications regulation, which may necessitate coordinated rather than disjointed approaches to regulatory intervention. The development of a formal interregulatory coordination framework, perhaps through the development of memoranda of understanding or regulatory instruments, may facilitate the exchange of information, the development of coordinated approaches to investigation, and the development of coordinated regulatory approaches to the complex jurisdictional issues that may arise in the context of the digital economy. The model established by the Digital Regulation Cooperation Forum in the United Kingdom, which coordinates the activities of the Competition and Markets Authority, the Information Commissioner’s Office, and Ofcom, provides an institutional framework that may be worthy of consideration in the context of the Indian regulatory reform process79.
CONCLUSION
The above discussion demonstrates the uncomfortable truth about the present state of digital competition enforcement in India: the legal framework governing the abuse of dominance was not designed with digital gatekeepers in mind. Section 4 of the Competition Act, 2002, is a tool that was designed in a world where pricing was a clear and present danger, and geographic market boundaries were sharply defined. But in the face of the fluid, multi-sided, and algorithmically mediated dominance of Google, Meta, and Apple, the very foundations of Section 4 begin to break down. The cases that are examined in this paper are reflective of the potential and limitations of the existing framework. The Google Android case highlighted the CCI’s capacity to apprehend ecosystem abuse and the leveraging of dominance in neighboring markets through tying and revenue-sharing agreements that are intended to deny access to competing search engines. However, the ₹1,337.76 crore penalty and remedial orders were issued after a delay of several years during which Google’s dominance was already established. This is directly relevant to the structural limitations of ex-post enforcement in dynamic markets. The WhatsApp privacy case extended the existing framework further by considering the mandatory imposition of expansive data-sharing obligations on users without alternative as an unfair condition of service under Section 4(2)(a) of the Competition Act. In this case, the Commission also made a tentative recognition of the potential for competitive harm in zero-price markets through degradation of privacy rather than price increases—an important conceptual development of some doctrinal significance. The Apple App Store case is still in its infancy and raises similar issues of aftermarket exploitation and tying that have yet to be fully interrogated by the existing framework.
What brings these proceedings together is the common problem of proof, and in this case, the problem is the inability of the regulators to access the logic behind the proprietary algorithms that govern the behavior in question. The exclusion resulting from search engine ranking manipulation, app store demotion, and interoperability degradation does not create a price-cost paradigm amenable to detection under the Areeda-Turner test. Network effects and data-driven lock-in mean that by the time the regulator is able to find abuse, tipping has already occurred. The zero-price model divorces the consumer welfare test from any reference to price, and the Commission is left to make inferences. These are not merely operational problems, but structural ones, and they cannot be addressed by more vigorous application of the existing framework.
The Digital Competition Bill, 2024, indeed reflects a sincere legislative effort to address the aforementioned gaps. The Bill, which proposes the imposition of ex-ante obligations on SSDEs with respect to self-preferencing, anti-steering, tying, and data intermixing without the need to establish harm, demonstrates a resolute move from adjudicative to structural regulation. The Bill follows the model of the European Union’s Digital Markets Act and is guided by the recommendations set forth in the Committee on Digital Competition Law Report (2023). The Bill recognizes the reality that the preceding analysis has conclusively demonstrated: the temporal disconnect between exclusionary conduct and remedial relief simply cannot keep pace with the speed at which digital dominance accumulates. Yet, the Bill as it presently reads generates legitimate concerns with respect to the imprecision of the SSDE designation criteria, the relative insubstantiality of the procedural mechanisms for contestation, and the absence of a methodology for assessing proportionality which potentially subjects the Bill to constitutional scrutiny under Articles 14 and 19(1)(g) of the Constitution of India.
Reform, therefore, must operate on two levels simultaneously. First, at the doctrinal level, Section 4 of the Act demands an ecosystem-based approach to market definition, moving beyond the artificial compartmentalization of each service offered by platforms. The predatory pricing test must be reconceptualized for a world of zero-priced services, and the evidentiary standard for algorithmic self-preferencing must be fleshed out more specifically, ensuring that any future proceedings are grounded more securely. Second, at the institutional level, what is urgently needed is for the Competition Commission of India to acquire dedicated technical expertise – algorithmic engineers, data scientists, and economists of platforms – without whom the ex-ante obligations under the Digital Competition Bill risk becoming mere theatre. Relevant here is the institutional investment made by the European Commission in its Digital Markets Unit before the DMA’s enforcement obligations came into play.
Digital competition regulation in India is at an inflection point. The jurisprudence built up through Google Android, WhatsApp, and Apple App Store cases is clear: algorithmic abuse is certainly within the legal cognizance of Section 4 of the Act – what is less clear is whether it is within its remedial reach. The Digital Competition Bill is designed to provide the structural tools to bridge the gap between legal recognition and effective enforcement, but this will depend on the sharpening of the criteria for designation, the strengthening of procedures, and the resourcing of the institutional foundations. The real challenge, as this paper has argued throughout, is not merely a legislative but an epistemological challenge for the regulator, who must find a way to grasp, audit, and evaluate technologies that are designed to be impenetrable. The way in which India navigates this challenge, between the need to regulate and the need to make sense of what is to be regulated, will, in no small measure, define the nature of competition in India’s digital economy in the decades to come.
*****
Footnotes
1. Jean Tirole, “Platform Competition in Two-Sided Markets,” 1 Journal of the European Economic Association 990, 991 (2003).
2. European Commission, Proposal for a Regulation on Contestable and Fair Markets in the Digital Sector (Digital Markets Act), COM(2020) 842 final, at 1–3 (Dec. 15, 2020).
3. Lina M. Khan, “Amazon’s Antitrust Paradox,” 126 Yale Law Journal 710, 743–47 (2017).
4. Organisation for Economic Co-operation and Development, Abuse of Dominance in Digital Markets 7–12 (2020). 5 Competition Act, 2002, § 4, No. 12, Acts of Parliament, 2003 (India).
5. Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 222–24 (1993).
6. Phillip Areeda & Donald F. Turner, Predatory Pricing and Related Practices Under Section 2 of the Sherman Act, 88 Harv. L. Rev. 697, 700–12 (1975).
7. Organisation for Economic Co-operation and Development, Algorithms and Collusion: Competition Policy in the Digital Age 7–9 (2017).
8. David S. Evans & Richard Schmalensee, Matchmakers: The New Economics of Multisided Platforms 211–14 (Harvard Business Review Press 2016).
9. Ariel Ezrachi & Maurice E. Stucke, Virtual Competition: The Promise and Perils of the Algorithm-Driven Economy 39–45 (Harvard Univ. Press 2016).
10. Competition Act, 2002, § 3, No. 12, Acts of Parliament, 2003 (India).
11. Lina M. Khan, Amazon’s Antitrust Paradox, 126 Yale L.J. 710, 716–19 (2017).
12. Organisation for Economic Co-operation and Development, Abuse of Dominance in Digital Markets 19–22 (2020).
13. Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on Contestable and Fair Markets in the Digital Sector (Digital Markets Act), art. 3, 2022 O.J. (L 265) 1, 12–14.
14. Jean-Charles Rochet & Jean Tirole, Platform Competition in Two-Sided Markets, 1 J. Eur. Econ. Ass’n 990, 992–95 (2003).
15. Digital Markets Act, art. 3(2).
16. Competition Comm’n of India, In Re: Allegations of Anti-Competitive Conduct by Amazon Seller Services Pvt. Ltd., Case No. 40 of 2019 (Jan. 13, 2020).
17. Jacques Crémer, Yves-Alexandre de Montjoye & Heike Schweitzer, Competition Policy for the Digital Era 65–68 (European Commission 2019).
18. Competition Act, 2002, § 4(2)(e), No. 12, Acts of Parliament, 2003 (India).
19. Carl Shapiro & Hal R. Varian, Information Rules: A Strategic Guide to the Network Economy 175–77 (Harvard Business School Press 1999).
20. Organisation for Economic Co-operation and Development, Network Effects and Competition Law: Note by the Secretariat 6–8, DAF/COMP(2018)2 (2018).
21. Crémer, de Montjoye & Schweitzer, at 34–38.
22. Lina M. Khan, Amazon’s Antitrust Paradox, 126 Yale L.J. 710, 780–85(2017).
23. David S. Evans & Richard Schmalensee, Matchmakers: The New Economics of Multisided Platforms 185-90 (Harvard Business Review Press 2016)
24. Ministry of Corporate Affairs, Government of India, Report of the Committee on Digital Competition Law 41–45 (2023).
25. Competition Act, 2002, § 4(1), No. 12, Acts of Parliament, 2003 (India). 37 Id. § 4, Explanation (a).
26. Id. § 19(4).
27. Id. § 4(2)(a)–(e)
28. Competition Comm’n of India, In Re: Alleged Cartelization in the Supply of Electric Power Steering Systems, Case No. 03 of 2014, ¶ 19 (Aug. 3, 2017).
29. In Re: Allegations of Anti-Competitive Practices by Google LLC, Case No. 39 of 2018 (Competition Comm’n of India Oct. 20, 2022), SCC OnLine CCI 93.
30. Id. ¶¶ 97–143.
31. Id. ¶¶ 210–228.
32. Id. ¶¶ 368–371.
33. Google LLC v. Competition Comm’n of India, Civil Appeal No. 8950 of 2022 (Supreme Court of India 2023), SCC OnLine SC 1419.
34. In Re: Updated Terms of Service and Privacy Policy for WhatsApp Users, Case No. 01 of 2021 (Competition Comm’n of India Apr. 24, 2021), SCC OnLine CCI 21. 47 Id. ¶¶ 44–52.
35. In Re: Complaints/References Against Apple Inc., Case No. 24 of 2021 (Competition Comm’n of India 2021); see also Regulation (EU) 2022/1925, art. 5(7), 2022 O.J. (L 265) 1 (prohibiting anti-steering obligations on gatekeeper platforms).
36. Einer Elhauge, Tying, Bundled Discounts, and the Death of the Single Monopoly Profit Theory, 123 Harv. L. Rev. 397, 402–08 (2009).
37. Areeda & Turner, supra note 13, at 700–12.
38. Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 224–26 (1993).
39. Competition Act, 2002, § 4, Explanation (b), No. 12, Acts of Parliament, 2003 (India).
40. Rochet & Tirole, supra note 25 at 1000–03.
41. Competition Comm’n of India (Determination of Cost of Production) Regulations, 2009, No. 3/2009 (India).
42. Organisation for Economic Co-operation and Development, Abuse of Dominance in Digital Markets, at 24–27.
43. Ministry of Corporate Affairs, Government of India, Report of the Committee on Digital Competition Law 38–42 (2023) [hereinafter CDCL Report].
44. Id. at 44–47
45. Id. at 51–56; see also Competition Law Review, Report of the Competition Law Review Committee 72–75 (Ministry of Corporate Affairs, Government of India 2019).
46. Digital Competition Bill, 2024, § 3, cl. 2 (India) (draft). Committee
47. Id. § 3(3)(a)–(f).
48. Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on Contestable and Fair Markets in the Digital Sector (Digital Markets Act), art. 3(2), 2022 O.J. (L 265) 1, 12–14 [hereinafter DMA]. 62 Id. art. 3(5)–(6).
49. Digital Competition Bill, 2024, §§ 9–14 (India) (draft).
50. Suzanne Scotchmer, Innovation and Incentives 27–32 (MIT Press 2004).
51. CDCL Report, supra note 56, at 63
52. India Const. arts. 14, 19(1)(g); see State of West Bengal v. Anwar Ali Sarkar, AIR 1952 SC 75, ¶ 18 (establishing the reasonable classification test under Article 14).
53. See Case T-684/22, ByteDance Ltd. v. European Commission (General Court of the European Union, pending) (examining proportionality of DMA gatekeeper designation).
54. DMA, supra note 61, recital 1–10.
55. Id. art. 3(2)(a)–(b).
56. European Commission, Commission Designates Six Gatekeepers under the Digital Markets Act, Press Release IP/23/4328 (Sept. 6, 2023).
57. DMA, supra note 61, art. 5(1)–(9).
58. Id. art. 6(1)–(12).
59. Id. art. 26(1)–(3).
60. European Commission, Digital Markets Act: Commission Opens Non-Compliance Proceedings Against Alphabet, Apple and Meta, Press Release IP/24/1689 (Mar. 25, 2024).
61. American Innovation and Choice Online Act, S. 2992, 117th Cong. §§ 2–3 (2022).
62. United States v. Grinnell Corp., 384 U.S. 563, 570–71 (1966).
63. Herbert Hovenkamp, Antitrust and Platform Monopoly, 130 Yale L.J. 1952, 1976–82 (2021).
64. Frank Pasquale, The Black Box Society: The Secret Algorithms That Control Money and Information 3–8 (Harvard Univ. Press 2015).
65. Crémer, de Montjoye & Schweitzer, at 112–16.
66. Competition Act, 2002, §§ 41–42, No. 12, Acts of Parliament, 2003 (India).
67. CDCL Report, supra note 56, at 89–92.
68. DMA, supra note 69, art. 26(2).
69. Competition Comm’n of India, Annual Report 2022–23, at 18–21 (2023).
70. Oles Andriychuk, The Dialectical Antitrust: A Philosophical Inquiry into the Nature of Competition Law 142–47 (Edward Elgar Publishing 2017).
71. Areeda & Turner, supra note 2 (Chapter II), at 700–12; Competition Act, 2002, § 4, Explanation (b), No. 12, Acts of Parliament, 2003 (India).
72. In Re: Allegations of Anti-Competitive Practices by Google LLC, Case No. 39 of 2018, ¶¶ 184–228 (Competition Comm’n of India Oct. 20, 2022), SCC OnLine CCI 93.
73. Competition Act, 2002, § 27, No. 12, Acts of Parliament, 2003 (India).
74. Id. §§ 36(2), 41(1)–(3).
75. Maneka Gandhi v. Union of India, (1978) 1 SCC 248, ¶¶ 56–58 (establishing the expansive content of procedural due process under Article 21 of the Constitution of India).
76. Crémer, de Montjoye & Schweitzer, supra note 79, at 50–56.
77. Competition Act, 2002, § 64, No. 12, Acts of Parliament, 2003 (India). 92 CDCL Report, supra note 1 (Chapter IV), at 71–74.
78. Competition Comm’n of India, Annual Report 2022–23, at 18–21
79. Digital Regulation Cooperation Forum, DRCF Workplan 2023/24, at 4–7 (2023), https://www.drcf.org.uk/workplan-202324.