Introduction: A New Era of Search or Control?

Once the reliable index of the open web, Google's search engine is currently going through a significant transformation. Its first international market rollout, AI Mode, was introduced in India in January 2025. AI Mode offers direct conversational responses produced by Google's massive language model, Gemini 2.5, in contrast to conventional search engines that present a list of links. The business also introduced AI Overviews earlier this year, essentially summaries that pull information from outside sources and show it to users directly. The implications are chilling for the rest of the internet, but the convenience for users is clear.

According to a July 2025 report by The Indian Express, users shown AI Overviews clicked through to traditional websites only 8% of the time, compared to 15% when the summaries weren’t shown. Even worse, clicks on the actual sources cited within AI summaries were as low as 1%. These statistics signal not innovation alonebut intermediation and displacement. When Google provides answers, it disincentivizes engagement with original content creators, many of whom depend on the referral topic for their survival.

The Competition Commission of India (CCI), India’s competition regulator, has historically been active in addressing Google’s dominance. But AI-driven changes in the search ecosystem present novel challenges, particularly around search bias, market foreclosure, and algorithmic self-preferencing.

The Shift from Search Engines to Answer Engines

From being a neutral gateway to becoming the destination itself, Google's most recent evolution represents an intellectual shift. Controlling not only where users go but also what they know is a new kind of market power brought about by this change.

Users can enter natural language queries and get context-aware, multi-paragraph responses when they are in AI Mode. Although they are still available, traditional organic results are now secondary, if not completely unnecessary. According to Google, this enhances the user experience. However, the action has come under fire from content platforms, publishers, SEO specialists, and more and more legal academics.

Data scraped from pre-existing content platforms, such as Wikipedia, news websites, product listings, and educational portals, is a major component of AI Overviews and AI Mode. But these original sources are rarely given proper credit. Web traffic and content production are thus separated, which could lead to the collapse of the digital publishing value chain.

India: The Ideal Test bed and the Perfect Storm

With more than 750 million internet users, India is Google's second-largest market. The majority of these users access the internet primarily through mobile devices. Additionally, voice-based search and Google Lens adoption rates are among the highest in the world in India.

The legal and regulatory environment of the nation is constantly changing in the digital sphere. The government's intention to strike a balance between innovation and accountability is indicated by the Digital Personal Data Protection Act, 2023, and the forthcoming Digital Competition Bill.

In light of this, Google has decided to launch AI Mode globally in India. However, traffic to Indian content platforms is already sharply declining, especially in verticals like e-commerce, news, education, and travel. This has an impact on user discovery, content longevity, and ad revenues.

Self-Preferencing in the AI Era: A Subtler Threat

In its 2018 decision in Matrimony.com v. Google, the CCI had already accused Google of self-preferencing in the days before artificial intelligence searches, and it had fined the company ₹1.36 billion. In that instance, the Commission determined that Google had unfairly favored its own services such as Google Flights over rivals in the search results. AI Overviews and AI Mode go one step further today. Google is curating the entire response rather than just showcasing its own services. It frequently integrates its vertical products, such as YouTube, Google Maps, and Google Flights, while marginalising rival platforms.

This isn't just algorithmic preference; it's AI-assisted user intent capture that leaves rivals with no options. The AI-generated response serves as the gatekeeper and the gateway.

Legal Framework: Abuse of Dominance under the Competition Act, 2002

India’s Competition Act defines abuse of dominant position under Section 4. Key provisions relevant to Goggle’s AI initiatives include:

- Section 4(2)(a): Prohibits unfair or discriminatory conditions imposed by a dominant enterprise.

- Section 4(2)(c): Targets practices leading to denial of market access.

- Section 4(2)(e): Addresses leveraging dominance in one market to enter or protect another.

All three clauses may be implicated by AI Overviews and AI Mode. They unfairly deny market access by monopolising users' attention, unfairly deny other websites referral traffic, and use Google's dominance in general search to further solidify its hold on the knowledge, travel, e-commerce, education, and health verticals.

Precedents Set by the CCI Against Google

The CCI has driven certain complex material driven precedents in the recent past:

- Search Bias Case (2018): In Matrimony.com Ltd. v. Google, the CCI ruled that Google had abused its dominant position by promoting its own verticals in search results. The ruling acknowledged that even algorithmically driven placements could amount to anti-competitive favourites.

- Android/MADA Case (2019–2023): Here, Google was found guilty of leveraging dominance in Android OS to promote Chrome and Google Search via restrictive agreements with OEMs. A penalty of ₹1,337.76 crore was imposed.

- Play Store Billing (2022): The CCI fined Google ₹936 crore for forcing app developers to use its proprietary billing system, curbing third-party payment gateways. The Commission also imposed behavioral remedies requiring Google to open up billing options.

Together, these cases signal that the CCI recognizes the multi-layered nature of dominance in digital ecosystems, and will not hesitate to act where market structures are distorted.

How the CCI Could Respond to AI Search Manipulation

As part of its ongoing scrutiny of artificial intelligence and digital markets, the Competition Commission of India (CCI) is already conducting a market study on “AI and Competition”, anticipated to conclude by the end of 2025. This study could serve as the foundational basis for formal enforcement actions against dominant tech platforms if competitive harm is identified—particularly in cases involving AI-driven search services. In the context of Google’s AI Overviews and AI Mode, the CCI has several statutory tools and precedents at its disposal:

1. Director General (DG) Investigation Under Section 26(1)

Should the market study reveal credible evidence of anti-competitive harm, the Commission may direct the Office of the Director General to initiate a detailed investigation under Section 26(1) of the Competition Act, 2002. Such an investigation could examine a range of exclusionary effects caused by Google's AI services, including:

The systemic decline in discoverability and visibility of third-party websites, especially in verticals like news, education, and commerce;

Preferential integration of Google’s own verticals (e.g., Google Flights, Google Shopping) into AI-generated results;

The opaque nature of AI algorithms determining which sources are summarized, cited, or suppressed—raising concerns over lack of algorithmic transparency and auditability.

2. Monetary Penalties Under Section 27(b)

If abuse of dominance is established, the CCI can impose monetary penalties under Section 27(b). These fines can be calculated as a percentage of the turnover relevant to the affected market. In the case of Google’s AI offerings, this would likely include revenue from search-based advertising, AI products, and affiliated vertical services, potentially amounting to significant sums given Google’s scale in India.

3. Behavioral Remedies for Market Correction

While financial penalties act as a deterrent, behavioural remedies are often more effective in correcting ongoing distortions in digital markets. In the context of AI search, the CCI could mandate Google to:

Clearly label AI-generated content as distinct from traditional search results to reduce user confusion and establish accountability;

Include prominent and clickable citations to all original source material used in AI Overviews, thereby restoring traffic pathways to publishers and service providers;

Provide opt-out mechanisms for content creators who prefer not to have their material processed or displayed by Google’s AI engines without explicit consent or compensation;

Ensure equal treatment of competing services in AI responses by applying non-discriminatory standards, thereby reducing the risk of subtle self-preferencing.

4. Structural Remedies in Case of Continued Non-Compliance

The CCI may look into structural remedies if it is determined that Google continues to use its dominance in spite of financial and behavioural interventions; however, such measures are only considered in extreme circumstances. To ensure that AI outputs do not unjustly benefit the company's other verticals, one potential solution would be to functionally separate Google's core search business from its AI-driven content summarisation units. This strategy would be similar to newly proposed antitrust laws in the US and the EU, where authorities are looking into comparable measures to reduce the monopolistic spillover effects of AI.

The Bigger Picture: Why This Case Matters

Google's shift to an AI-first platform signifies the privatisation of information discovery and goes beyond a simple product change. The danger of informational monopolies arises as AI technologies start to shape how billions of users obtain news, information and products. For a country like India, where digital infrastructure is crucial to inclusive growth, this raises broader questions around algorithmic accountability, data ethics, and platform neutrality.

Conclusion

Google's introduction of AI Overviews and AI Mode represents a strategic realignment of digital power rather than just a technical advancement. Google runs the risk of making the same antitrust mistakes it has in the past, but with a smarter interface, by blocking external links, integrating its own services, and providing no transparency into the curation of AI answers. The CCI of India has the legal framework, precedent, and resources to react. The objective must always be the same, whether it be through sanctions, corrections, or structural change: to maintain a competitive, open, and diverse digital ecosystem. The law must make sure that AI doesn't become a gatekeeper for speech, commerce, or knowledge as it becomes the new internet interface.

Aditya Maurya is a LL.M. student at Gujarat National Law University, Gandhinagar.