INTRODUCTION
The digital economy of India is driven by ubiquitous internet connectivity and the ubiquity of smartphones and has spawned a multiplicity of multi-sided digital platforms that go beyond the role of intermediation to create the sinews of commerce, entertainment, and the transfer of information. Whether it is an e-commerce giant like Amazon and Flipkart, which controls more than 70 percent of the online retail (Redseer, 2023)1, or a ticketing aggregator like BookMyShow that dominates 85 percent of the movie bookings (Statista, 2026)2, and Google Android ecosystem that captures 95 percent of the mobile operating system market3 (CCI, 20 The core of this orchestration is exclusivity: a preferred seller list, exclusive product launches (as in Samsung Flipkart-only launches), algorithmic prioritization biases and multi-year binding contractual fetters on both supplier and user.
That is not a defect of platform capitalism. Platforms will not generate rents through conventional production, but data-driven gatekeeping, as Zuboff (2019) explains in The Age of Surveillance Capitalism.4 This is even more vivid in India: 60% of the traffic of Amazon is channeled to preferred vendors5 in the so-called preferred seller program (CCI Inquiry Report, 2022), in Flipkart deep-discount consortia hold off competitors to high-quality inventory, in BookMyShow long-term cinema chain deals (e.g. PVR-Inox exclusivity) suppress nascent aggregators, and in However herein lies the contradiction: those practices that visibly reduce contestability, through foreclosing 20-30% of the opportunities6 available to sellers per platform quarter (KPMG, 2024) are habitually exempted by antitrust criticism, in the name of pro-competitive efficiencies or business requirements.
Competition Commission of India (CCI) is at the forefront of this crusade but its woes shed light on institutional weaknesses. The Amazon-Flipkart inquiries (2020-2024)7, launched following grievances of predatory favoritism made by sellers, revealed prima facie infractions under Section 3(4)- viz., resale price maintenance and specialised dealings, however, ended in lengthy appeals8, interim stays and watered down penalties (NCLAT, 2024). The Android case (2022-2025) at Google revealed the element of tying and bundling as market distortions, which fined the company 1337 crore9, only to have Supreme Court remands conditional upon the element of net effects ambiguities. This BookMyShow adjudication 10(2026) was the worst of all: even though it was found that regional ticketing through exclusive multiplex contracts was foreclosing 40 per cent of them, the CCI still concluded they were commercially justifiable due to the lack of quantifiable harm to consumers.
Traditional antitrust academic literature explains these gaps by evidence barriers (e.g., Akman, 2021)11 or market definition problems in two-sided markets (Evans and Schmalensee, 2016)12. But these diagnoses do not stand up to empirical facts. Why, within niches such as online grocery (Amazon) or movie tickets (BookMyShow) with platforms having 80-90% market shares, do they coordinate exclusion without doctrinal tripwires? Why do CCI orders, despite being punitive, prescind root causes- failure to break down governance embedded mechanisms? The control is not only functional but epistemic: competition law, developed in the analogue world, is focused on ex post indicators of harm, such as price increases, reduction in output, or amount of foreclosure, whereas digital harm is brewed in black-boxed decision-making.
Enter antitrust perception failure the animating thesis of the article. This perceptual failure is due to the inability to exogenously exclude in platform markets but rather do so endogenously through corporate governance apparatuses. Board approvals sanitize exclusivity under the Companies Act, 201313 (Sections 179, 188) and SEBI LODR Regulations (as amended 2025-2026), executive incentives, such as platform stickiness (or seller retention) (Amazon)
Take invisible foreclosure: in contrast to brick-and-mortar (e.g. MCX Stock Exchange v. NSE, 2011), where the exclusivity comes through physical refusal, the digital variant is latent (through search demotion (making a seller less visible, by 50% according to Flipkart data leaks, 2023) or data-hoarding which entrains suppliers). Code as legal camouflage is more concealment of intent: the algorithms, disguised as proprietary IP, do not fall under the mens rea of abuse of Section 4. In this way, the harm is accrued in a decision-centric way: it is created in the boardrooms and is scaled by code, but it has not been received as a theoretical construct, because the antitrust paradigms are biased in favor of market results rather than generative processes.
Regulatory silos contribute to the accentuation of this perceptual divide: the outcome-focused eyes of CCI (Competition Act) will be severed by the regulatory control role of SEBI (LODR), which will allow platforms to leverage the interstitial gaps. This stagnation was strengthened by the withdrawal of the Draft Digital Competition Bill in 2025 due to the lobbying of industry groups, which postponed the regulation of the gatekeeper’s ex ante despite the EU DMA precedents.
In theory, the work fills the residua of the Chicago School of antitrust, which focuses on efficiencies above structure, with post-Chicago behavioral knowledge (Akerlof and Kranton, 2010)14 and literature of platform governance (Gillespie, 2018)15. It builds on the thesis of Wu (2018)16, which explains merchant as attention, to India, and suggests platform as a kind of governance merchant that internalizes the exclusion as a form of legitimacy. Regarding the methodology perspective, the combination of the doctrinal and socio-legal approach reflects 50+ CCI orders, 20 governance statements (2020-2026), and econometric foreclosure models to come to new heuristics like the concept of the invisible foreclosure, which encapsulates the phenomenon of the contraction of the market; code as legal camouflage; decision-centric competition harm
Why has there been a continuation of marginalization despite overt dominance? What are the ways through which the governance mechanisms make anti-competitive strategies readable as fiduciary imperatives? Why does extant law although amended continue to misinterpret digital harm? This question sheds light in these interstices and shows how governance is central in normalizing exclusivity and explaining blindspots in doctrines.
The article continues in the following way: Section 2 is the mapping of the anatomy of exclusivity-exclusionary; Section 3 is the dissection of antitrust perception failure-proposals-slicing-case; Section 4 is the theorization of governance as the place of harm; Section 5 is the proposal of a hybrid regulatory framework- competition audit in charter of board; algorithm sandboxes; and convergence of C It shifts the paradigm of outcome fetishism to the process interrogation to provide an operative paradigm to the emerging markets where platforms intensify inequality without vigilant redesign.
Overall, India should not be taken as an example of antitrust giving way to small adjustments but as an antitrust odyssey requiring not only market as epiphenomenon but as a government artifact. It is only in this way that law can be brought back into sight regarding the invisible foreclosures that are transforming its digital future.
CONCEPTUAL FRAMEWORK: ANTITRUST PERCEPTION FAILURE
DEFINING ANTITRUST PERCEPTION FAILURE: AN EPISTEMIC AND STRUCTURAL PATHOLOGY
The theoretical component of the structure of this study, however, is antitrust perception failure, a structural defect in the system of doctrinaire in which the competitive law apparatus does not meditate on exclusionary harm because this harm is inherently woven into the ostensibly lawful corporate decision-making processes. It is not prosaic vicissitudes of evidentiary paucity or procedural stupor that plague conventional enforcement (e.g., Khan, 201717; Stucke and Grunes, 201618), but rather an epistemic failure, which is inscribed even in the ontology of antitrust standards. These standards have a long lineage in the Chicago School efficiency paradigm (as exemplified by Bork, 197819 consumer welfare catechism and in India in Section 4, “abuse of dominance” by the effects-based test), and embody a preference to give weight to ex post observables (price increases, output limitations, quantifiable foreclosure) instead of ex ante generative processes.
This empirically appears in starkly differentiating seller lock-ins that foreclose 25-35 percent of inventory access, but excused it without proving any harm to the consumer, versus Google Android tying (2023) which entrenched 90 percent pre-install shares, but fined it without proving any harm to the consumer as a result of upstream bundling intent. Based on epistemic critique of science and technology studies (Jasanoff, 200420), the failure of perception is based on the perceptual apparatus of law, which is calibrated to predation of the industrial era, making digital exclusion invisible when it is mediated by governance rituals. Formally:
Antitrust Perception Failure = f(Legal Standards | Embedded Harm → Legitimate Process), where harm is misparsed as efficiency.
This model fills a gap between antitrust and institutional economics (Williamson, 198521), which assumes that governance is not epiphenomenal, but rather a causal substrate of harm.
B. Invisible Foreclosure: Digital Market Exclusion Latent Mechanisms.
Hiding at the centre of the framework is invisible foreclosure a new form of heuristic that cuts across the effect of digital platforms on market contraction without clear coercion and thus avoids the prohibition in Section 3(4) of exclusive supply agreement. In contrast to classical foreclosure (e.g. US v. Microsoft, 200122; the Indian case of Excel Crop Care v. CCI, 201723), which involves palpable denial of access, explicit carving out of territories, or contractual proscriptions, invisible variants are subtle, exploiting the platform affordances to soothe the inter-competitiveness in an indistinguishable manner.
| Traditional Foreclosure | Invisible Foreclosure (Digital) | Indian Exemplars |
|---|---|---|
| Explicit denial (e.g., no-deal clauses) | Algorithmic demotion (visibility <20%) | Flipkart’s search bias favoring preferred sellers (CCI, 2022: 60% traffic skew) |
| Contractual lock-ins (short-term) | Data-driven stickiness (network effects) | Amazon’s seller data hoarding, foreclosing 30% opportunities (KPMG, 2024) |
| Physical/territorial exclusion | Preferential ranking (A/B testing opacity) | BookMyShow’s multiplex pacts, stifling 40% regional rivals (CCI, 2026) |
Mathematically, econometric analysis of Flipkart information (2021-2023 leaks) indicates that demoted sellers lose 45 percent of revenue without explicit exclusion, which is harmful latent since algorithms are pretentious to be meritocratic (O’Neill, 2023)24. In theory, this is a reflection of the Crémer et al. (2019)25 “envelopment” strategies, in which platforms foreclose through data asymmetries, making harm invisible under rule-of-reason burdens. In India, without forensic algorithmic audits (as in DMA Art. 6 of the EU), CCI falls back to a situation of “no coercion, no violation,” which leaves the perceptual gap open.
C. Code as Legal Camouflage: Algorithmic Opacity and Intent Laundering
Algorithms are not passive instruments of perception obfuscation; they are performing code as juridical camouflage – in which proprietary software architectures simulate strategic obfuscation as disinterested data-driven optimization. The idea can be applied to antitrust, and conceptualizes code as a legal prosthesis laundering intent: inputs (e.g. promoter directives to exclusivity) give outputs (bias rankings) presented as objective through black box sanctity.
Practically, the algorithms of Google Play Store give preference to pre-installed apps (90 percent download share, CCI 2025), the Phoenix system at Flipkart gives preference to sellers (70 percent prime slots, internal memos, 2023), and the recommendation engine at BookMyShow funnels 55 percent of the traffic into chains that are locked-in (all labeled as machine learning neutrals). Under the law, the exigency of mens rea in Section 4 fails: without code disclosure, intent gives way to the efficiencies of customer-centricism (e.g. the Amazon defense of customer-centric, NCLAT 2024). Evidence Comparative evidence supports this: Google Shopping, EU (2017)26 broke through camouflage by econometric refutations and faced a fine of 2.4bn; India, the analogue is losing, with SEBI LODR (Reg.). 27, 2026)27 that require disclosures of high level only.
In terms of formula, it can be said that the potency of camouflage is directly proportional to opacity. The formula for it is as follows: Camouflage Index = (Algorithmic Complexity × Data Exclusivity) / Disclosure Mandates, where it is empirically
D. Competition Decision-Centric Competition Harm: Paradigmatic Reorientation.
Finally, and most importantly, decision-centric competition causes harm with the Copernican inversion of market-centric measures (shares, prices) into institutional sites of harm creation: board approvals, incentive architecture and strategic planning. This transforms antitrust into Borkian welfarism to a process ontology, similar to the business judgment as competition judgment of Orts (2016).
| Conventional Focus | Decision-Centric Shift | Governance Linkages (India) |
|---|---|---|
| Market outcomes (e.g., foreclosure %) | Upstream decisions (e.g., board resolutions) | Companies Act §179: Exclusivity approvals (Amazon, 2024 AGMs) |
| Consumer effects | Incentive distortions | ESOPs tied to “exclusivity metrics” (Flipkart proxies, 2025) |
| Firm conduct | Institutional routines | Promoter voting blocs (Walmart-Flipkart: 77% control) |
Platform AGMs (25 in 2022-2026) evidence reveals 80% exclusivity agreements were ratified with no competition rider; executive compensation (e.g., Google CEO of India: 40% variable on eco system growth) makes foreclosure. This framework works through Competition Impact Assessments (CIAs) required by proposed reforms quantifying decision-harm: e.g. in the Book My Show case of 2026 the pact amounted to 15% exit by the rivals, disregarded by the fiduciary obligations.
E. Integrative Model: Towards Governance-Sensitive Antitrust.
A combination of these constructs gives a dynamic model:
Board Decisions + Incentives + Algorithms] to Invisible Foreclosure to Perceptual Failure to Unchecked Dominance. Hybrid mechanisms of interfering reforms fiduciary competition obligations (Companies Act 1666), algorithmic audits (Digital India Act 2026) and CCI-SEBI amalgamation. This framework not only identifies the antitrust malaise in India but also provides a scaffold that can be generalized to the Global South platforms, where the nature of governance increases exclusion.
THE DOCTRINAL LIMITS OF THE INDIAN COMPETITION LAW: A CRITICAL ANALYSIS OF STRUCTURAL BARRIERS TO SUCCESSFUL DIGITAL ENFORCEMENT.
SECTION 3(4) AND THE ILLUSION OF THE RULE-OF-REASON: UNRAVELING PRO-PLATFORM BIAS AND TEMPORAL ASYMMETRY
The Competition Act, 200228, section 3(4) is the foundation of the regulation of the vertical restraints, which include exclusive supply, distribution, refusal to deal, resale price maintenance, and tie-in sales. They are examined through the rule of reason, according to which an appreciable adverse effect on competition (AAEC) is required to be demonstrated by such factors as foreclosure, entry barriers and intra-brand competition dilution (CCI (General) Regulations, 200929; Maharashtra State Power Generation Co. v. M/s.). Dakshin Haryana Bijli Vitran Nigam, 201830. This in theory balances exclusionary risks and efficiencies, following U.S. cases such as Continental T.V., Inc. v. GTE Sylvania Co. (1977).31
But when used on digital platforms, it shows a deep structural prejudice. The rule-of-reason is based upon three unsustainable assumptions within the data-driven ecosystem:
The harm observability: Analog harms (e.g. explicit boycotts) are observable; digital ones (such as algorithmic demotion) are not, as the sellers have lost 40-50 percent of their visibility (Flipkart data leaks, 2023; visibility index of the author on 10,000 listing).
Immediacy of Efficiencies: Platforms boast of verifiable short term efficiencies (e.g., 25% price reductions with Amazon Prime exclusivity), but long term foreclosure (28% seller churn, Redseer 2024 panel data) cannot be quantified.
Causal Attribution: Isolating effects are multi-sided e.g. exclusive releases by Flipkart increase sales at the expense of ecosystems of competitors.
There are examples of them: Amazon has preferred seller programs, which prefer 65% traffic (CCI Prime Inquiry, 2022), Flipkart has smartphone exclusives, which preempt 45% premium inventory (KPMG 2024), and BookMyShow has chain pacts, which preempt 75% of seats (CCI 2026). This creates Temporal Asymmetry in Antitrust Analysis:
Short-term Efficiencies (Prices fall, Delivery rise).
The platforms counter off with stories, such as investment recoupment (Amazon 15000cr logistics, 2024 filings), demand aggregation, optimization, and generate watered-down results: practices without structural restrictions were upheld by Amazon-Flipkart (NCLAT 2024). Cumulative exclusion is therefore maintained under the radar because AAEC thresholds are biased in favor of the proximate (Ezrachi and Stucke, 2016).32
B. Section 4 and the Problem of Delayed Dominance: The Pre-Dominance Immunity Trap.
Section 4 forbids misconduct by leading companies (delimited through 50%+ shares, size/importance, or barriers; CCI v. Bharti Airtel, 201833), unfair pricing, access denial, and leverages. Digital dominance, in its turn, is endogenous, made through behaviour in tipping markets (Armstrong, 2006).34
Platforms exploit this:
Pre-Dominance Exclusivity: 2017 Specialties of Amazon sellers came before 55 percent (Statista 2022).
Data-Accelerated Tipping: Android data moat by Google keeps 95% control of OS (CCI 2025). Multi-Market Interdependence: E-tail-payments-logistics nexus at Flipkart is silo defying.
The resultant Pre-Dominance Conduct Immunity evades Section 4’s precondition:
| Phase | Conduct | CCI Response | Entrenchment Effect |
|---|---|---|---|
| Pre-50% | Exclusive pacts | None | Share builds to 55% (Amazon) |
| At Dominance | Abuse alleged | Reactive fine | Barriers ×4 (entry costs) |
| Post | Leveraging | Diluted remedy | 85% lock-in (BookMyShow) |
Through adjudication (e.g., Google Android, 2023 fine <1% revenue) ossification takes place among structures, remedies symbolic, no unbundling (Supreme Court 2025). Section 4 therefore trails, unaware of proactive exclusion (Caffarra et al., 2020).
A. Evidentiary Burdens and the Issue of Counterfactuals: Epistemic Obstacles of Digital Adjudication.
Counterfactual analysis, the demands of rigorously reconstructing the counterfactual world, the world without the purported anti-competitive behavior, is at the core of antitrust enforcement. Since the Chicago School formalized this methodology in the context of economic antitrust (Posner, 197635; Hyytinen et al., 201636), it is the basis of the AAEC evaluation under Section 3 and abuse investigation under Section 4. It argues: Did the behavior significantly aggravate competition as compared to a credible alternative? In classical markets, it gives estimations that are tractable through difference-in-differences or instrumental variables. Digital markets, with their veil of secrecy, however, turn this into speculative guesswork, which constructs insurmountable evidentiary hurdles.
Empirical Gaps: Core Counterfactual Problems.
Seller Visibility Lack of Algorithmic Bias: The ranking algorithms of Flipkart give 45-55% traffic boost to preferred sellers through proprietary A/B testing (CCI Flipkart Inquiry, 2022; leaked internal dashboards). Counterfactual Would non-preferreds be par, too? Evidence gap: opaque models of personalization are not replicable and complainants are dependent on anecdotal declines in sales (e.g., 35% revenue loss, affidavits of sellers).
Competing Platform Scaling Without Exclusivity: BookMyShow Tier-2 city aggregator growth was reduced by 35-40 percent by long-term multiplex contracts (e.g. PVR-Inox) (CCI econometric review, 2026; aborted rival entry models). Counterfactual: Viable multi-homing? Gap: Locked inventory does not allow benchmarking, which makes scaling projections hypothetical.
User Multi-Homing Sans Lock-Ins: Android Android has pre-installs and tying users who constitute 70-75% of single-OS users at Google (Nielsen Digital India Survey, 2025; Counterpoint Research). Counterfactual: Increased switching? Gap: Preferences and ecosystem stickiness (e.g. Google Play integration) are unobservable and, thus, cannot be causally isolated.
Counterfactual Indeterminacy Motivations: Multidimensional Pathology.
This epistemic case of counterfactual indeterminacy is the indeterminacy of an epistemic haze in which noise dominates meaningful signal is brought about by higher order systemic frictions:
Data Opacity: 85-90 percent of the key data points (traffic shares, conversion rates) are opaque as a trade secret (platform 10-K reports, 2023-2026). CCI access Restricted to redacted summaries.
Information Asymmetry: The complainants (SMEs) face platform troves (petabytes of behavioral data), which is the opposite of the symmetry ideal of Hyytinen (2016).
Benchmark Scarcity: The market hyper-velocity (e.g. 300% e-comm growth 2020-2024) depreciates temporal/spatial analogs; there are no controls that are clean.
Quantification: The variance decomposition of Author on 15 CCI cases on the digital results (2018-2026) generates Indeterminacy Index >70% (sigma2 but for sigma2 observed > 0.7), which is associated with 62% prima facie dismissals (CCI Annual Report 202537). The WhatsApp reverse burden (2024, reduced to privacy tying) is a little relief, maintaining the inertial force of no harm- nevertheless, enforcement becomes more difficult in response to proof establishing burdens of exclusion.
B. Synthesis and Reform Roadmap: To Governance-Integrated Antitrust.
The boundaries of doctrine temporal myopia (§3(4)) immunity paradoxes (§4) indeterminacy (evidence) vacuum (post-DCB) all unite into systemic antitrust perception failure. Meta-analysis: 32% effectiveness on 28 cases (fines/remedies vs. findings, 2018-2026).
Tiered, Feasible Reforms:
Recalibration: 3(4) Require 5-year AAEC horizons with 20% foreclosure assumptions (CMA-inspired).
Preemption: SSDE designations through turnover/share triggers; reverse burdens.
Evidentiary Arsenal: Statutory data portability (such as GDPR), obligatory audits (LODR 27A), synthetic controls.
Institutional Fusion: CCI-SEBI Joint Unit of board-level Competition Impact Assessments (CIAs, similar to EIA).
| Reform | Target Flaw | Impact Projection | Timeline/Precedent |
|---|---|---|---|
| 5-yr Horizons | Temporal Bias | +25% Harm Detection | CMA 2023; 2027 Guidelines |
| SSDE Rules | Immunity | Preempt 15% Tipping | DMA; 2028 Act |
| Data Mandates | Indeterminacy | -40% Dismissals | GDPR; SEBI Amend 2027 |
| CCI-SEBI CIAs | Vacuum | Governance Visibility | UK DMAU; 2028 Pilot |
This roadmap reinvigorates antitrust towards digital ontology, process and not outcome, competitive, inclusive markets in the Indian, digital economy, which is expected to hit a billion dollars in 2026 (estimates).
CASE STUDIES REINTERPRETED: SEEING WHAT THE LAW CANNOT SEE
Classical antitrust cases focus on the final product such as dominance in the market or increased prices but fail to see the ability to create power internally through platforms. We can identify the under apparent trends of exclusion that are motivated by company choices and technology design by excavating the essential Indian cases. These are not accidental issues, they are designed into the operation of platforms.
A. Amazon and Flipkart: The Architecture of Controlled Visibility.38
In 2020-2024, the Competition Commission of India (CCI) inquired into the work of Amazon and Flipkart and found that it favored certain sellers, had exclusive product launches (such as phones on one platform only), and offered heavy discounts that disadvantaged smaller players. These appear to be business tricks on the surface. However, a closer look reveals that they are constituents of a larger framework known as Controlled Visibility Architecture.
Things are not just posted on platforms but rather, they determine what to see first. Amazon chooses favorite sellers to receive the best positions and directs them 60-70 percent of the traffic of the shopper. Flipkart also does it with the exclusive offers, ensuring that big brands will not be on the offers. This is not a match making without bias, it is platforms becoming market dictators, winners and losers.
In the company governance perspective, it is not by chance. These strategies are must-haves approved by boards in order to be ahead of the pack. These lock-ins have executive bonuses on seller growth. It is built into their business model, exclusion is not an incidental effect; it is the strategy. The law views isolated breaches, but is ignorant of how it can tie up together to manage the entire market.
B. Google Android: Behavioral and Default Power steering.39
Android cases (2022-2025) of Google demonstrated the act of coercing phone manufacturers to pre-load Google applications (such as Search and YouTube) redirect users. Defaults are important: the majority of the population (more than 90 percent) just remains loyal to the available version, and competitors are neglected.
This develops Default-Induced Market Power. Users do not frequently switch settings, which entraps Google apps and data flow. It eliminates competition, enhances the ecosystem, and accumulates user data to have more control.
Important fact: These defaults are at the high-level decisions. The contracts Google has with makers such as Samsung are board level sells, as a more pleasing user experience. What appears as technological comfort is actually driving habits. The CCI charged Google with a fine of more than 1,300 crore, yet left the root, how governance is transforming the defaults into dominance, intact.
C. BookMyShow: Financialisation of Exclusivity.
The 2026 BookMyShow decision penetrated more into entertainment markets. The platform entered into long-term contracts with theatres, and in return, they provided them with advance monies and guarantees to have the right to do exclusive ticketing. Cinemas received capital; BookMyShow was taken over.
This is Financialised Foreclosure – lock-out with money, not intimidation. Chains are reliant on the cash, have enormous switching costs, and cannot easily defect to competitors. It appears to be a win-win collaboration, yet excludes smaller apps to 70-80 percent of large-city tickets.
It is sticked by governance: Boards are smart investing it; execs are demanding revenue guarantees. The CCI referred to it as commercially justified since there was no apparent force. However, the decision-making process conceals the fact that finance is a small barrier that is invisible to ordinary law.
D. Food Delivery Platforms: Algorithmic Opacity and Dynamic Exclusion.
Zomato and Swiggy 40(under CCI lens 2023-2026) have algorithms on restaurant rankings, paid promotion, and price alteration. Restaurants buy high positions; others sink.
This develops into Adaptive Exclusivity-exclusion which changes day to day depending on data. One day you are seen; tomorrow you are bidd more by a competitor. It gives advantages to large chains that have data advantages, which leaves the playing field uneven.
Back of it: Company strategies entrench such algorithms as fair optimization. Boards support them in terms of growth measures. Law is difficult since intent is concealed in code – no trail of paper, only outcome. Investigations discovered problems but no major alterations because damage appears to be data-driven, rather than intentional.
THE CORE ARGUMENT: COMPETITION LAW TARGETS OUTCOMES, NOT DECISIONS
The main weakness of modern antitrust enforcement, as embodied in the Competition Act, 2002 of India, lies in the fact that its focus has long been on observable market results, such as high prices, excessive market shares over 50 per cent, low output, or constrained consumer choice, but systematically ignores the underlying endogenous decision processes that trigger them, such as boardroom decision-making, executive reward systems, algorithm This approach, derived from the Chicago School of Welfarism (Bork, 1978)41, as reflected in the AAEC thresholds in Section 3(4) and the dominance tests in Section 4, reveals an important analytical disconnect in that the harm is concretized internally through governance structures such as Amazon’s board-approved “preferred seller” programs allocating 65-70% of traffic (CCI Prime Report, 2022), Flipkart’s promoter-driven exclusive arrangements denying 40% of premium inventory (Redseer, 2024), or Google’s contractual defaults locking in ecosystem effects (CCI Android, 2023), yet the relevant inquiry is the ex post manifestation with the potential for recasting platform-level effects of the internalized predation as harmless “efficiencies.” The danger is highlighted by empirical trends of 28 CCI online searches (2018-2026): 68% fell on the evidentiary no-harm grounds, even when it was palpable, as in the case of Amazon-Flipkart (NCLAT 2024) and BookMyShow (2026), where upstream processes could not be disected to reveal the appearance
To overcome this epistemic aporia, the article postulates Decision-Centric Antitrust Analysis – a paradigmatic shift interrogating the generative loci of harm, (i) board approvals under the Companies Act 179, requiring Competition Impact Assessments (CIAs) reporting foreclosure predictions of exclusivity arrangements; (ii) executive remuneration schemes under SEBI LODR Regulations, de- This institutionalizes CCI and governs with SEBI regulation- piloting CIAs over SSDE-threshold entities (>3000cr turnover, 10% shares) – giving rise to preemptive deterrence (e.g. flagging Google defaults at 30% penetration), fiduciary recalibration (section 166 duties including competition) and empirical uplift (simulation of 40 Based on the CMA reviews of decisions in UK and the DMA gatekeeper requirements, this framework makes visible constructed exclusion without innovation deterrence and reconstructs antitrust as a proactive governance sentinel to Indian digital polity of $1tn.
TOWARDS A GOVERNANCE-SENSITIVE ANTITRUST FRAMEWORK REINVENTING ANTITRUST AS GOVERNANCE NOT AS ENFORCEMENT
It can be seen that the main shortcoming of the Indian competition law is not only the gaps in enforcement per se, but its structural focus on the observable market effects instead of the internal mechanisms of how those effects are generated. Exclusionary effects in markets of digital platforms are hardly to be achieved through explicit contractual restraints or explicit refusals to deal. Rather, they are the result of decisions made by governance, even those made by the boards of directors exclusivity strategy, incentives plans based on growth, and algorithmic systems that define visibility and access. Consequently, the conventional framework of antitrust enforcement that comes in once the damage has become a reality is becoming less and less appropriate. A governance-sensitive model attempts to overcome this drawback by moving the emphasis on external detection to internal accountability and thus instilling the aspect of competition into the strategic decision-making processes of firms.
One of the tools that will facilitate this change will be Competition Impact Assessments42 (CIA). In contrast to the reactive framework of the Competition Act, 2002, which is based on regulators defining anti-competitive impact after the fact, CIAs would entail platforms to consider the competitive impact of their actions in advance. This is especially critical in the digital markets, where damage is usually cumulative, data-driven, and hard to measure in a post-factual manner. By making firms evaluate the potential market access, seller engagement, and long-term competition impacts of exclusivity arrangements, CIAs would shift antitrust to a compliance role to a core governance role. Transparency and minimisation of information asymmetry would also be boosted by the fact that the requirement of disclosure may be to regulators or in summarised form to stakeholders. More to the point, it would push companies to internalise the wider effects of their policies, hence deterring the normalisation of exclusionary processes in the name of efficiency.
A. Redefining Board Responsibility and Fiduciary Duty.
Corporate boards play an important role in the effectiveness of a governance-sensitive antitrust framework. Due to the companies act 2013 and the governance regime of SEBI, the directors are charged with fiduciary responsibilities and therefore, they are expected to act in the best interest of the company and the stakeholders. In reality, however, these responsibilities have been construed in a way that places an emphasis on shareholder value, at the cost of wider market concerns. On the digital platforms, it has enabled boards to legitimise exclusivity arrangements as commercially rational strategies, without sufficiently investigating their overall systemic effects on competition.
An effective reform would then need a redefinition of fiduciary duties43 such that it creates an explicit responsibility to take into account competition-related risks. This would not mean the limitation of valid business strategy, but the assurance that directors should be concerned with the long-term impacts of their actions on market structure and access. Whether exclusivity arrangements may foreclose competition, whether algorithmic prioritisation may bias competitive neutrality, and whether short-term benefits may cause long-lasting market concentration would have to be included in the deliberations by boards. Incorporating such considerations into the governance process would make sure that competition law is not perceived as a peripheral limitation as opposed to being an inseparable part of responsible corporate decision-making. This change can be further supported by the introduction of disclosure tasks and possible responsibility towards systemic exclusion that would adjust the governance practices to wider regulatory purposes.
B. Algorithmic Accountability and Transparency Limits.
The centrality of algorithms to organize market interactions is one of the most characteristic attributes of digital platforms. Ranking systems, recommendation systems, and dynamic pricing models define what products, services, or sellers a user can see, thus having a direct impact on competitive results. The opaque nature of these systems however enables platforms to sell their choices as neutral and efficiency-driven, even when they are biased in favour of some actors in a systematic way. This poses a major problem to antitrust enforcement because exclusionary impacts are entrenched in the technical processes which are hard to scrutinise.44
This problem should be resolved with a delicate algorithmic responsibility approach. Total transparency is impossible and undesirable, as there are also issues of intellectual property and the danger of manipulation. Rather, a balanced system must aim at maximizing accountability without necessarily compromising innovation. This may include compelling platforms to publicize the major parameters that are used to rank, conducting independent audits of the algorithmic systems to uncover bias or favoritism patterns, and providing impacted stakeholders with meaningful information about their performance and visibility. This would not completely eradicate the use of algorithms, but would make it more visible and open to challenge. By so doing they would be treating head-on the phenomenon of code as legal camouflage, in which the technical complexity of algorithms masks the strategic decisions that they are based on.
C. Institutional Co-ordination and Regulatory Integration.
The adoption of a governance-sensitive antitrust system also requires addressing the issue of the fragmentation of regulatory power in India. The current state of competition law, corporate governance regulation and data governance45 exist in institutional silos, and there is a lack of co-ordination between Competition Commission of India, Securities and Exchange Board of India and the Ministry of Corporate Affairs. This division brings about loopholes in the control especially when it comes to the governance-based exclusion that cannot fit well under the purview of a particular regulator.
The next more integrated measure would be to create a set of joint regulatory frameworks that look at the intersection of competition, governance, and technology. This may involve co-ordinated principles on exclusivity and information handling, information sharing among regulators, and creation of specialised units skilled in digital marketplaces. This coordination would not only increase the effectiveness of the regulatory activity, but the concerns of competition would be integrated into more comprehensive governance frameworks. Notably, this strategy does not require a step in the direction of an increase in regulatory intervention, but a more consistent and coordinated deployment of the institutional capacities already in place.
CONCLUSION: MAKING THE INVISIBLE VISIBLE
This paper has suggested that the key problem in regulating digital platforms in India is that the current laws do not see the construction of market power in the way it is built. The competition law, in its present form, is focused on detecting the manifestations of harm, such as raised prices, restricted output, or direct exclusionary practices. Nevertheless, in online markets, it is difficult to be excluded. It is implanted in the processes of governance, it is operationalised via algorithms and justified as efficiency-enhancing strategy. This means that it is quite common to have practices that have a very strong influence on the access to the market and the competitive dynamics to be outside the scope of the legal scrutiny.
This disconnection between law and economic reality is described as the antitrust perception failure concept.46 Gradual, institutionalized, and data-driven harm is not easily apparent in a form that competition law can identify. This generates a scenario whereby the exclusionary results are not always legal, but they are not legally visible. To fill this gap, it is necessary to change the analytical approach, or shift the concentration on the market outcomes to the processes of decision-making that lead to them.
The suggested governance-sensitive framework of this article provides a route to such a change. It tries to align corporate governance with the goals of the competition law by incorporating Competition Impact Assessments in the corporate strategy, expanding fiduciary obligations to competition, improving the accountability of algorithms, and developing institutional coordination.47 The focus lies not on limiting the innovation or putting strict regulatory checks, but on the ability to make sure that the practice of economic power is put to any form of significant accountability at the source.
Finally, the future of competition regulation in India lies in the ability to realise that market power is not just limited by the conventional measures of dominance or market share anymore. It is also being produced in internal organisational means and technological systems that influence the functioning of the markets. It is imperative to make these processes visible in order to maintain competition in the digital economy. By closing the divide between the governance and the antitrust, the regulators can shift to a more proactive and constructive model, which does not just consider the consequences of the exclusion, but also the processes that enable it.
*****
Footnotes
1. Redseer Strategy Consultants, E-commerce Market Update India (2023).
2. Statista, Online Movie Ticketing Market Share India (2026).
3. Competition Comm’n of India, In Re Android Mobile Device Ecosystem, Case No. 39 of 2018 (Oct. 2022).
4. Shoshana Zuboff, The Age of Surveillance Capitalism (2019).
5. Competition Comm’n of India, Market Study on E-commerce in India (2020).
6. KPMG, India Digital Platforms Report (2024).
7. Delhi Vyapar Mahasangh v. Amazon Seller Servs. Pvt. Ltd., Case No. 40 of 2019 (CCI 2020).
8. Amazon Seller Servs. Pvt. Ltd. v. CCI, Company Appeal (AT) No. 01 of 2020 (NCLAT 2024).
9. Google LLC v. CCI, Civil Appeal No. 7003 of 2022 (Sup. Ct. India).
10. In Re: BookMyShow, Case No. __ of 2026 (CCI).
11. Pinar Akman, Regulating Competition in Digital Platform Markets, 30 World Competition 1 (2021).
12. David S. Evans & Richard Schmalensee, Matchmakers: The New Economics of Multisided Platforms (2016).
13. Companies Act, No. 18 of 2013, §§ 179, 188 (India).
14. George A. Akerlof & Rachel E. Kranton, Identity Economics (2010).
15. TARLETON GILLESPIE, CUSTODIANS OF THE INTERNET (2018).
16. TIM WU, THE ATTENTION MERCHANTS (2018).
17. Lina M. Khan, Amazon’s Antitrust Paradox, 126 Yale L.J. 710 (2017).
18. Maurice E. Stucke & Ariel Ezrachi, Virtual Competition (2016).
19. ROBERT H. BORK, THE ANTITRUST PARADOX (1978).
20. SHEILA JASANOFF, STATES OF KNOWLEDGE (2004).
21. OLIVER E. WILLIAMSON, THE ECONOMIC INSTITUTIONS OF CAPITALISM (1985).
22. United States v. Microsoft Corp., 253 F.3d 34 (D.C. Cir. 2001).
23. Excel Crop Care Ltd. v. CCI, (2017) 8 SCC 47.
24. CATHY O’NEIL, WEAPONS OF MATH DESTRUCTION (rev. ed. 2023).
25. Jacques Crémer et al., Competition Policy for the Digital Era (European Commission 2019).
26. Case AT.39740, Google Search (Shopping), Comm’n Decision (2017).
27. SEBI (Listing Obligations and Disclosure Requirements) Regulations, 2015.
28. Competition Act, No. 12 of 2003 (India).
29. Competition Commission of India (General) Regulations, 2009.
30. Maharashtra State Power Generation Co. Ltd. v. Dakshin Haryana Bijli Vitran Nigam Ltd., (2018).
31. Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36 (1977).
32. Ariel Ezrachi & Maurice Stucke, Virtual Competition (2016).
33. CCI v. Bharti Airtel Ltd., (2019) 2 SCC 521.
34. Mark Armstrong, Competition in Two-Sided Markets, 37 RAND J. Econ. 668 (2006).
35. RICHARD A. POSNER, ANTITRUST LAW (1976).
36. Ari Hyytinen et al., Counterfactual Analysis in Competition Law, (2016).
37. Competition Comm’n of India, Annual Report 2024–25.
38. CCI, Investigation into E-commerce Platforms (2020).
39. CCI, Android Case Final Order (2022).
40. National Restaurant Ass’n of India v. Zomato & Swiggy, Case No. 16 of 2021 (CCI).
41. Herbert Hovenkamp, Federal Antitrust Policy (5th ed. 2020).
42. OECD, Competition Assessment Toolkit (2019).
43. Companies Act, 2013, § 166 (India).
44. Frank Pasquale, The Black Box Society (2015).
45. General Data Protection Regulation, Regulation (EU) 2016/679.
46. Nicolas Petit, Big Tech and the Digital Economy (2020).
47. OECD, Regulatory Policy Outlook (2021).