This piece is originally published on Global Competition review at Data & Antitrust Guide- Third Edition.

This article outlines the differences between the Indian regulatory environment in the AI driven markets as opposed to digital markets in two separate sections.

AI regulation and governance

The Indian economy is one of the fastest growing in the world and the Indian government has set an ambitious target of becoming a US$5 trillion economy by 2027. AI adoption and development are key drivers in supporting this target, with India being one of world’s largest and fastest growing AI markets, projected to reach US$131.31 billion by 2032 with a growth rate of at a compound annual growth rate of 42.2%.^[1] The growth in AI adoption in India is evenly spread across sectors such as finance, healthcare and manufacturing. Local players are offering AI-enabled products and solutions in abundance, for example, Uniphore (conversational AI and business automation) and 5C Network (AI-powered radiology),^[2] among others. The potential for growth in AI is also attracting significant investments in this space by global firms like Microsoft (US$17.5 billion),^[3] Google (US$15 billion)^[4] and Amazon (US$35 billion).^[5] The nascency of the AI ecosystem and the larger aim to increase AI adoption poses the challenge for the government of achieving a fine balance between regulation and pushing for immense innovation potential. Thus far, the government has been cautious in regulating AI, and the focus appears to be on fostering a more trustworthy AI ecosystem.

That said, discussions around regulation of AI have intensified over recent months. Questions regarding the need for regulation and the shape and form of said regulation have been prominent in both government and policy discussions. This importance was highlighted when India hosted the International AI Summit in early 2026. The summit saw participation from over 100 countries, 22 heads of states and more than 60 regulatory bodies and government dignitaries discussing the broader contours of AI risks, governance and regulation. A consistent takeaway was that at this stage regulatory efforts should be primarily directed towards ensuring AI systems are safe, reliable and resistant to misuse. Antitrust intervention does not appear to be an immediate priority; regulators should instead focus on developing a nuanced understanding of the technology and market dynamics as they evolve, and only then assess whether, and in what form, competition regulation may become necessary.

AI governance in India

AI governance in India is primarily led by the Ministry of Electronics and Information Technology (MeitY), with the Competition Commission of India (CCI) focusing only on anticompetitive conduct in AI-driven markets. The MeitY governs issues related to information technology, focusing on safe, ethical and responsible AI adoption. It issued the AI Governance Guidelines in November 2025,^[6] which were endorsed by 22 countries,^[7] explicitly favouring “innovation over restraint”. The Indian AI governance architecture centres on: trust; openness; accountability; and inclusive development.^[8] The goal is to ensure that AI models are reliable, secure and trustworthy, thereby building consumer trust to enable increased AI adoption by consumers and institutions, creating a positive feedback loop that helps grow domestic AI enablers and downstream players. A significant part of the regulatory effort has been dedicated to explaining potential concerns to stakeholders, ensuring the industry is aware of its responsibilities as the market expands. Interestingly, during the AI summit, India achieved a Guinness World Record for the “Most pledges received for an AI responsibility campaign in 24 hours”, with over 250,000 validated pledges reaffirming public commitment towards responsible AI adoption.^[9]

The CCI, on the other hand, is tasked with regulating anticompetitive conduct. As part of its advocacy mandate, the CCI published its AI Market Study Report in October 2025.^[10] The report identifies potential risks, including algorithmic collusion, anticompetitive conduct (eg, bundling and self-preferencing), discriminatory pricing, entry barriers, reduced transparency and market concentration. These risks are noted across sectors such as e-commerce, healthcare, financial services and logistics. The report suggests self-regulation is an important tool to keep markets competitive and suggests an AI toolkit for self-regulation based on principles of accountability, documentation and risk mitigation, as summarised below.

CCI’s recommended toolkit for AI deployers

The CCI’s AI toolkit is a governance blueprint, not a technical manual. It requires companies to know how their AI systems work, anticipate how those systems might distort market outcomes and build internal accountability before and after deployment. In effect, the CCI is signalling that responsibility for preventing anticompetitive outcomes sits with enterprises designing and deploying AI, and that a failure to understand or monitor AI behaviour will not be an acceptable defence. Some key recommendations to keep in mind if businesses are deploying AI include:

• for algorithms, maintain internal clarity (documentation and explainability of design logic and outcomes) and external disclosure (inform users about AI interaction and identifying factors influencing promoted results for platform neutrality). Another suggestion is to test in sandbox environments (ie, conduct controlled experiments in different scenarios before rolling out the algorithm to the public and establish triggers for human review of algorithmic decisions);
• collect only data necessary for the AI’s stated purpose;
• regularly audit AI outputs for unintended biases, predatory pricing patterns or discriminatory treatment of specific consumer segments;
• educate data scientists and product managers on competition law to ensure “compliance by design” from the earliest stages of AI development; and
• systematically evaluate AI systems for risk and assign clear compliance responsibilities, ensuring senior management oversight for high-risk deployments and implementing an impact-based approval process for market-affecting AI.

The toolkit sets out recommendations for AI deployers, including a framework for self-audit, an implementation process and a suggested self-audit checklist, along with suggested documentation to be maintained by AI deployers. Enterprises enabling or providing AI based services are expected to conduct self-assessments based on the self-audit checklist.^[11] If these are demonstrably met, the CCI is unlikely to intervene, and there has been no active enforcement to date.

CCI’s approach to ex post facto and ex ante AI regulation

At the AI summit, Chairperson Ravneet Kaur noted that there is interest in potentially taking action in the future, signalling a “wait-and-see” approach.^[12] In parallel, the CCI is closely monitoring acquisitions and partnerships across the AI value chain for potential concentration risks.^[13] It is also exploring international cooperation frameworks to facilitate knowledge sharing and collaborative policy development, reflecting the inherently cross-border nature of AI markets.^[14] For now, the CCI’s approach is to clearly articulate risks while encouraging responsible self-regulation. While immediate intervention appears unlikely, this position remains dynamic and will need to be closely monitored as markets develop.

Concerning ex ante regulation, there is no clarity on how the CCI is likely to treat AI, but the CCI Chairperson has stated on multiple public forums that the AI Market Study has concluded that AI-driven markets mirror risks similar to digital markets.^[15] Keeping pace with global developments, including the enforcement of the EU Digital Markets Act (DMA) in 2024, discussions around ex ante regulation of digital markets gained momentum in India. The Ministry of Corporate Affairs (MCA), which is the governing ministry of the CCI, released a comprehensive draft Digital Competition Bill (DCB) in March 2025 after extensive consultations with relevant stakeholders. The release of the draft DCB was followed by extensive stakeholder consultations that concluded in July 2025. Subsequently, in August 2025, the MCA issued a report^[16] on the CCI’s evolving role in new technology markets and the implementation of an ex ante framework (Report). The Report identifies the concentration of economic power among large tech platforms, which leverage their market positions to entrench dominance and expand into adjacent markets. In this context, the Report recommended that the DCB must:

• be enacted in a phased, evidence-based manner with a focus on pre-emptive regulation of “gatekeepers” or “systemically significant digital enterprises”, aligned with the enforcement mechanisms under the DMA;
• introduce rebuttal mechanisms and specific designation thresholds to avoid capturing growing Indian firms; and
• ensure that these thresholds are carefully managed and implemented based on the findings of a formal market study.

Apart from the existing digital services covered in the DCB (search engines, browsers, social media, etc),^[17] the Report suggests including virtual assistants (eg, Siri and Alexa) under DCB and flags agentic AI (systems that can act on their own) as a likely CCI priority. However, it adds that the DCB should avoid blanket prohibitions of specific conduct, allowing for context-specific assessments. While the draft DCB primarily targets gatekeeper platforms in digital markets, its final form remains under review. It will be interesting to see if and how AI-specific anticompetitive concerns, such as data monopolisation for training models, algorithmic tying and cloud compute dominance, will be integrated into the DCB’s ex ante framework if and when the bill is adopted, or if AI will end up with a separate regulatory mechanism. Moreover, the India-EU Free Trade Agreement^[18] could fortify the Indian government’s mandate to introduce ex ante digital markets regulation, but its approach to AI remains undecided. There have also been heightened discussions on AI regulations, including the March 2024 advisory^[19] on AI models by the MeitY and the India AI Governance Guidelines,^[20] amid global parallels like the EU AI Act.

Conclusion

The Indian government’s current approach reflects a broader ambition to grow across the entire AI stack, from foundational models and infrastructure – via a 103 trillion rupee IndiaAI Mission funding for 18,000 GPUs and sovereign datasets by 2026^[21] – to applications, and adoption of sovereign AI models,^[22] such as Sarvam AI and BharatGen (for languages), positioning the country as a global leader in ethical, homegrown AI. The 2025–2026 vista confirms a highly strategic regulatory forbearance regarding AI, designed to facilitate the growth of AI capital economy and understanding the technology before usage. While the CCI is tackling anticompetitive conduct in digital markets rapidly, underscored by massive fines and structural remedies, with AI it recognises that premature, heavy-handed intervention could stifle the investments needed to build a robust domestic AI ecosystem. The mandate remains to support the Indian economy by scaling AI safely, issue information regarding potential risks and rely on self-regulation.

A differential approach in digital markets

Unlike AI, enforcement in digital markets remained robust in 2025, accounting for almost 21% of all the prima facie orders and final orders passed by the CCI. Between 2014 and 2024, there were 99 cases involving digital markets before the CCI, with 64 dismissals.^[23] Another key development in this space is the CCI’s first ever settlement under the newly introduced commitment and settlement regime,^[24] where it closed a five-year long investigation into Google’s licensing agreements for Android TVs.^[25] The CCI was primarily investigating Google for imposing restrictive agreements on its partner original equipment manufacturers (OEMs) that mandated bundling of Google apps with the Play Store (a necessary app according to the CCI). The CCI accepted Google’s proposals to offer a licence for standalone Play Store on Android TVs without other Google apps and to waive off certain compatibility requirements. The settlement process took a little less than a year and the CCI accepted the proposals on payment of a settlement fee of 202.4 million rupees (approximately US$2.1 million). The CCI’s order is precedent-setting for future commitment and settlement processes.

Enforcement trends in key sectors

In 2025, antitrust enforcement in digital markets experienced significant activity, especially in the operating system, app store, online retail, e-commerce, online advertising, online messaging services and fast moving consumer goods (FMCG) sectors.

Pragmatic and evidence-based evaluations

There is no automatic penalisation of dominant tech companies and the CCI looks at evidence of market foreclosure. For instance, the CCI dismissed abuse of dominance claims against Microsoft,^[26] Google,^[27] BookMyShow,^[28] Zomato^[29] and Rapido,^[30] because it found their practices to be commercially justified, optional or standard industry practice without sufficient evidence of market denial. Relevant details of these decisions are set out below:

• The Microsoft Windows and Defender decision: the CCI dismissed a complaint alleging that Microsoft abused its dominance by tying its preinstalled Microsoft Defender antivirus software with Windows OS, and that Microsoft imposed one-sided anti-malware API licence and listing agreements, and mandated participation in the Microsoft Antivirus Initiative programme, forcing third-party developers to surrender proprietary technological information and competitive control to operate effectively on Windows OS. The CCI noted that the antivirus market remains dynamic and competitive. Based on this, and setting an evolved benchmark for treatment of theories of harm associated with preinstallation, the CCI held that, even if Defender is preinstalled, there is no compulsion on users to use it and users can freely install third-party options, automatically disabling Defender. It further noted that, though non-MVI applications cannot automatically disable Microsoft Defender, they can operate concurrently and can be distributed through Microsoft Store or downloaded directly from their websites.
• The Google Play Store Policy Enforcement decision: the CCI closed two separate complaints against Google filed by app developers Liberty Infospace and Zucol Solutions. Liberty Infospace alleged that Google terminated the app developer’s account without proper notice owing to its “relation ban” policy. Zucol’s entire developer account (and not just the allegedly violating app) was terminated after it attempted to bypass a prior ban. Zucol’s app (Pobreflix) was flagged for IP and security violations, and when the original account was banned, Zucol created a “related” account to continue operations. The CCI found no prima facie case of abuse of dominance in both instances. It concluded that, in Liberty Infospace, Google’s actions were a valid enforcement of its “relation ban” policy to protect users from malware, since the complainant had links to a previously violating account (Google linked Liberty’s account to its CTO, who had been associated with accounts flagged for distributing “malware seeds”). In Zucol, the CCI upheld Google’s ban at the level of developer’s account and not just the offending app, in a ruling that indicates that account level termination is a valid deterrence to repeat offenders and is not anticompetitive. The CCI also held that Google’s appeals process, which includes human intervention, was reasonable and consistent with industry practice.
• The BookMyShow decision: the CCI, while concluding BookMyShow holds a dominant market position in online intermediation services for booking movie tickets in India, ruled against allegations of abuse, including discriminatory practices and restrictive agreements with single-screen cinemas. The CCI dismissed the complaint, finding that practices like seat reservations and differential treatment of single-screen cinemas and multiplexes were objectively justified to prevent overlapping bookings, and unsustainable as the two cinemas in issue differed in terms of infrastructure, cost and service quality. Furthermore, the CCI reasoned that: data ownership exclusivity clauses are standard industry practice; cinemas collecting an advance payment from BookMyShow was operationally justified as it provided them with an interest free working capital; lock-ins were justified to provide BookMyShow sufficient time to recover the advance paid; and, most importantly, all of these conditions were part of a mutually negotiated contract and therefore not abusive. Lack of foreclosure due to the above was further substantiated by the presence of major rivals like Paytm and Amazon. Importantly, aligning with its previous orders, the CCI reiterated that it cannot rule on the fairness of the commission fee as it is not a price regulator.
• The Zomato decision: a consumer complaint alleged that Zomato engaged in unfair practices by charging higher prices than restaurant menus and imposing mandatory platform fees and default settings tipping delivery partners. The CCI found no prima facie violation, noting that the fees were not discriminatory and features like tips included easily visible opt-out methods.
• The Rapido decision (on-demand transportation services): the CCI recently dismissed a complaint against Rapido, an Indian ride-hailing company. The complaint alleged that Rapido abused its dominant position by providing services at a lower price, as compared to its competitors and by using private vehicles without necessary permits. The CCI held that the allegations fall within the purview of the Motor Vehicles Act 1988, a special legislation and therefore ousts the jurisdiction of the CCI.

Recognising privacy as a competition parameter

Recently, the court that hears CCI appeals, the National Company Law Appellate Tribunal (NCLAT), held in the Meta judgment^[31] that the CCI has jurisdiction to examine data practices that harm competition, confirming that data privacy is a key non-price parameter and that mandatory, unfair data-sharing terms (like the WhatsApp 2021 privacy policy) can be an abuse of a dominant position.

WhatsApp and Meta appealed a CCI order regarding WhatsApp’s 2021 privacy policy, that required users to accept expanded data-sharing with Meta without an opt-out option. In late 2025, the NCLAT partially upheld the CCI order, including the penalty of 2.13 billion rupees (approximately US$22.8 million) on WhatsApp and Meta for exploiting their dominant position through the policy, and allowing WhatsApp to share user data with other Meta companies or Meta products for advertising. The NCLAT held that WhatsApp is dominant in the market for over-the-top (OTT) messaging apps through smartphones in India, and sharing of data led to entry barriers or denied market access to Meta’s rivals in the display advertising market. However, the NCLAT dismissed the CCI’s finding that Meta leveraged its dominance in the OTT messaging apps through smartphones in the Indian market to maintain or increase its market presence in the display advertising market. It held that a claim for leveraging only survives if dominance in the first market is used by the “same enterprise” to enter or protect another market, and Meta and Whatsapp are separate legal entities despite one being the other’s parent.

Two appeals filed by Meta and WhatsApp before the Supreme Court, including an interim application seeking a stay on NCLAT’s order dated 4 November 2025, were dismissed, as the appellants agreed to comply with the directions of the NCLAT by 16 March 2026.^[32] The Court directed the appellants to file a compliance certificate before the CCI, and directed the CCI to file a response to the compliance certificate before it. The matter is pending before the Supreme Court.

Overturning preventive forward-looking remedies

In the Meta judgment (detailed above), the tribunal overturned the CCI’s blanket five-year ban on cross-data sharing with other Meta entities if there is user consent to share the data (consumers can opt-in or opt-out). Similarly in the GPBS judgment,^[33] the NCLAT set aside the CCI’s forward-looking directions on potential or likely harm.

In March 2025, the NCLAT upheld the CCI’s finding that Google abused its dominance by mandating the use of the Google Play Billing System (GPBS) and leveraging its app store dominance to protect its Unified Payment Interface payment app (Google Pay). The NCLAT reversed the CCI’s remedies, concluding that it prematurely treated Google as a “gatekeeper”. These overturned remedies included mandates for transparent data policies, prohibiting leveraging billing data for competitive advantage and requiring a fair, transparent fee structure. The NCLAT held that Indian competition law requires proof of ex-post violations, not pre-emptive gatekeeper regulations.

Separately, the tribunal reversed the findings of discriminatory pricing, restriction of technical development and denial of market access holding:

• Google’s policy of charging third-party developers on the Play Store a service fee while exempting its own revenue generating app, YouTube, was not discriminatory, and made commercial sense; and
• the CCI’s claim that Google making use of GPBS mandatory for app developers on Play Store impedes innovation and results in denial of market access lacked concrete evidence and ignores the fact that payments using Google’s billing system accounted for less than 1% of overall UPIs, leaving the rest open to competing payment service providers.

The NCLAT applied the principle of proportionality in levying penalties and remedies. In the GPBS judgment, the tribunal significantly reduced Google’s penalty from 9.36 billion rupees (US$100.19 million) to approximately 2.16 billion rupees (US$23 million) by roughly 75%, by calculating the penalty based strictly on “relevant turnover”, that is, Play Store-specific revenue in India. The NCLAT held that penalties must be tied to the relevant business and product under investigation, not a company’s entire footprint.

Positioning the CCI as a non-price regulator

The CCI has consistently held that it does not function as a price regulator, citing that companies are permitted to charge fees proportionate to their commercial realities, and permitting varying revenue models and fee structures if they reflect commercial realities and explicitly enable end users to opt-out (eg, BookMyShow decision and Zomato Decision).

In the Google ATV settlement – the CCI’s first, and to date only, settlement – the CCI declined to consider the “fairness” of fees charged for “Google’s New India Agreement”, a standalone Play Store licence without mandatory pre-installation of other Google apps like YouTube. The CCI noted Google’s submissions that the fee offsets lost revenue from not requiring pre-installation of Google apps, and original equipment manufacturers are open to generate revenue through promotional agreements with competing app developers for pre-installation and prominent placements.

Further sector updates

Operating Systems and App Store Ecosystem

This sector saw significant activity, particularly involving Microsoft (Microsoft Windows and Defender decision) and Google’s market practices (Google Play Store Policy Enforcement decision and the GPBS NCLAT judgment).

Online retail and e-commerce

Decisions in this sector, including the BookMyShow and Rapido decisions, focused on platform intermediation, exclusivity and pricing practices. A key ongoing investigation is outlined below.

Some micro, small and medium-sized traders of smartphones and related accessories had approached the CCI in 2019, alleging various vertical agreements between Amazon and its preferred sellers and Flipkart and its preferred sellers, both allegedly leading to foreclosure of all non-preferred sellers from these e-tailing marketplaces. They also alleged that these preferred sellers are in reality an extension of the two e-commerce players and are in effect controlled by the two. Their main concerns were:

• preferential treatment of a small group of sellers;
• deep discounts relating to certain categories of products, including exclusive deals with mobile phone manufacturers;
• predatory pricing; and
• circumventing foreign direct investment rules that apply to e-commerce in India.

The CCI issued prima facie orders directing investigation.^[34] Amazon and Flipkart separately challenged the prima facie orders against them. While Amazon’s challenge failed, Flipkart got a favourable order from the Supreme Court in February 2026, which set aside^[35] NCLAT’s findings,^[36] and remanded the matter back for fresh consideration as to whether there is a prima facie case for investigation that warrants further remand to the CCI.

Currently, the investigation report against Amazon has been submitted to the CCI and parties who were not part of the prima facie order but have been added to the investigation by the DG have challenged their addition without prior notice and an opportunity to defend before the Karnataka High Court. The High Court has stayed further proceedings until a final decision on this challenge.^[37]

Online advertising

The focus here remains heavily on Google’s ad-tech intermediation services.

Early in January 2025, the CCI received a complaint from an individual app developer, Maulik Surani,^[38] with allegations relating to ad tech services similar to the issues in the ongoing News Publishers investigation.^[39] The Alliance of Digital India Foundation (ADIF) also filed a complaint^[40] before the CCI against Google’s ad tech practices in the online search and display advertisement markets. The CCI has split ADIF’s complaint into three categories of allegations:

• Case No. 23(1) of 2024, in the online display advertising market, which the CCI has joined with the ongoing News Publishers investigation, given overlapping issues (ie, Google indulges in self-preferencing by: tying the services of its publisher ad server and supply side platform (SSP) (DFP) with Ad Exchange (AdX), and offering them as a single Google Ad Manager function; tying the services of its SSP (DV360) with AdX; and linking access to its vertical’s YouTube ad inventory to the use of its SSP DV360 and so on. The CCI has directed the DG to submit a consolidated report.
• Case No. 23(2) of 2024, in the search advertising market regarding Google Ads policies, which the CCI has dismissed because these allegations have already been addressed in substance in previous cases.^[41]
• Case No. 23(3) of 2024, on miscellaneous allegations (ie, removal of 3P cookies, leveraging search dominance to unfairly profit with search ads and opaque ad review and redressal).

Another key update in this sector is the publicly reported raid by the DG at the Indian officers of global advertising agencies like GroupM, Interpublic, Publicics and Dentsu, early in 2025.^[42] The current status of this investigation is unknown.

Online messaging services

The key decision in this sector in the past year is the Meta NCLAT judgment.

FMCG

Apart from the closure order in the Zomato decision, this sector is currently facing escalating scrutiny as listed below:

• Swiggy and Zomato: the CCI is investigating food delivery platforms Zomato and Swiggy following a complaint by the National Restaurant Association of India (NRAI) alleging anticompetitive practices such as preferential treatment, exclusivity and price parity. The CCI issued a prima facie order to examine issues regarding exclusivity, price parity and platform neutrality,^[43] directing its investigative arm, the DG, to conduct a detailed probe. Media reports^[44] indicate that, after an unusually long 30-month investigation, the DG has recommended both companies be found to be in violation of the Competition Act. There is no final decision.
• Quick-commerce antitrust complaint: the All India Consumer Products Distributors Federation (AICPDF) has reportedly filed a complaint with the CCI against quick-commerce giants Blinkit, Zepto and Swiggy Instamart. The federation alleges predatory pricing and deep discounting that create an unfair playing field for small offline retailers.^[45] There is no published prima facie order yet.
• Ultra-fast food delivery friction:the NRAI reportedly plans to intensify its complaints against Zomato and Swiggy, specifically concerning their new 10-to-15-minute food delivery apps, Bistro and Snacc.^[46]

Conclusion

The regulatory shift in India is increasingly evident. The CCI is adopting a more forward-looking approach, introducing preventive measures to govern the conduct of dominant digital platforms, rather than focusing solely on past behaviour. There is also a clear movement towards assessing the effects of a firm’s conduct, rather than relying only on its size or market position. Recent decisions reflect targeted enforcement and a deeper, more nuanced understanding of digital markets and platform dynamics. Notably, in 2026, closure orders have outnumbered investigation orders, indicating a more selective and evidence-based approach to enforcement. For businesses and policymakers, the challenge is no longer just tracking if a market firm is dominant, but how that dominance is exercised in a multi-sided ecosystem.

Endnotes