Introduction

In early 2023, the AI landscape crossed a defining threshold: OpenAI’s ChatGPT reached 100 million active users within just two months of launch, making it the fastest-growing consumer application in history. As of 2025, just a handful of firms – OpenAI (with Microsoft), Google DeepMind, Gemini, Anthropic & a few others – command the vast majority of generative Artificial Intelligence (AI) ecosystem, controlling not only the most powerful foundation models but also the underlying compute infrastructure & critical training datasets.

Despite this staggering growth & concentration, dominant AI players get the exception of belonging to “nascent market” to deflect stricter antitrust scrutiny. The argument goes: AI is new, unpredictable & evolving so rapidly that market power today may not be tomorrow. However, competition market regulators must not be relaxed by this “false comfort”. As seen in India’s own competition jurisprudence, where Competition Commission of India (CCI) has rejected “nascent market” defence stating that dominance can solidify swiftly, particularly in digital markets with strong network effects & high entry barriers.

This blog argues that in the age of AI, market definition must adapt, separating hype from structural reality & ensuring that claims of “nascency” do not become convenient cover for anti-competitive market. 

Defining Relevant Market in the Era of AI services with unclear boundaries

Emergence of nascent competition is a double-edged sword: on one hand it can increase innovation, improve quality, whereas if not addressed, it can lead to reduced innovation, decreased consumer choices & increased market concentration. The emerging technologies & innovation at an unprecedented pace in the realm of AI are leading to unclear boundaries, especially where two different models of AI are linked together. For example, the Large Language Models (LLM) in AI models are different from Cloud-Based AIs. 

It is well established in competition law that market definition needs to be considered in the context of the theory of harm in the case at hand & must be carried out with regard to the overall economic context. Historically, the Indian competition authority has defined relevant markets in a narrow manner.  In the M/S Ess Cee Securities Pvt. Ltd v. M/S Dlf Universal Limited, CCI defined the relevant market restrictively. To define relevant market, the nature of the market must be kept in mind & in a market characterised by continuous technical evolution, there is a high risk that if the market is defined broadly by encompassing wide range of technologies having minor differences. AI markets are both extremely broad, encompassing a range of different products and technologies and extremely concentrated.

A key aspect which influences the market definition & competition structure in AI evolution is the foundation models, which serve as a underlying framework upon which several AI applications are built into an AI stack. Foundation models may not be fully interchangeable, or interchangeable at all, depending on their training data, foundation models could be general purpose or domain specific. Thus, there is a need to define separate relevant markets      for different foundational models as there will be a range of potential business & functionalities with each different model, due to which if the relevant market is not defined narrowly, would lead to companies escaping liability for their anti-competitive behaviour. 

New Factors of Dominance Emerging with Tech Innovations in AI sector to be taken Account

It is well settled that holding a dominant position under Section 4 of the Competition Act 2002 (Competition Act) is not per se prohibited by law, but abusing the same is punishable under the law. In digital platform markets, the key role played by gatekeepers, network economies & tipping effects, combined with large tendency of digital market to enter into another often makes it difficult to distinguish between legitimate competition on the merits & abuse of dominance less obvious, particularly under the guise of nascency. Resultantly they are not being caught when abusing this dominant position by entering into exclusive agreements of providing data to train their LLMs models or tying in two separate products. In the MCX Stock Exchange Ltd. & Ors vs NSE Ltd., the CCI analyzed whether NSE took defence of Nascent market. However, the CCI rejected the defense & held that "Nascence" refers to the immediate existence after birth, while "infancy" is the time after. Markets are considered nascent during the first few months, followed by infancy for another year, then maturity when the market is fully developed. AI market has evolved in 2019s with peak in 2023s, thus this is not the nascency phase of this market.

NSE was held to be dominant due to its high degree of vertical integration, ranging from data information to high technologies information. Thus, “nascency” should never be the sole factor to be taken to assess dominant position in the market especially in cases of AI markets where there are dynamic tech revolutions going on.

Superior technology is considered an indicator of dominance in the market by the CCI in Shri Shamsher Kataria v. Honda Siel Cars India Ltd. Recently, Google’s DeepMind’s solution to the protein-folding problem through AlphaFold was a calculated decision that gave it a formidable edge in bioinformatics and pharmaceutical AI. In industries like biotech and healthcare AI, where DeepMind currently has a first-mover advantage, this is more than just technical superiority, but it is market capture.

Other factors include economic power of the enterprise, including commercial advantages over competitors, and market structure in size of the market. For example - OpenAI signs deals with news publishers, Reddit, Stack Overflow to use unique content for training because of its economic capacity to enter into such agreements. Competitors without these deals can’t match training data quality. Recently, Amazon dominates voice AI and cloud AI by sheer size. AWS holds 32% of the global cloud market, giving it unmatched scale to train massive AI models.

Data forms the foundation for AI as past behavioural patterns preserved in data are essential to train and increase the efficiency and accuracy of algorithms' predictions. This source data is combined with third-party data, and the aggregate is critical for improving existing AI, in terms of producing accurate behavioural data. Presently, OpenAI’s GPT models, trained on vast data with Microsoft’s Azure cloud, are far ahead in capabilities, an advantage startups cannot easily match. This tech advantage attracts huge investment, reinforcing dominance of GPT with $80 B. Huge size and advanced resources help the enterprise to have an advantage over other players and allow them to create barriers.

The CCI observed that ‘in a data-driven ecosystem’, unreasonable data collection and its sharing may grant undue advantage to dominant players in the market in the case of In Re Privacy Policy for WhatsApp Users. Similarly, Data collection and size of enterprise are directly proportionate, reinforcing enterprises' market power, as held in In Re: Delhi Vyapar Mahasangh and Flipkart Internet Private Limited. The massive volumes of data, combined with machine learning and AI capabilities, strengthens the market position of these data-driven technology enterprises. When transitioning to new platforms, a lack of data retention combined with a lack of interoperability creates a significant barrier to entry with high switching costs. Any new entrant therefore has to create disruptive technologies to compensate for the significant switching costs associated with attracting people from an established market. As a result of using this data, competitors will never be able to match the costs of competitive products released by tech giants. 

OpenAI’s best models run exclusively on Microsoft Azure, no other cloud provider can host them, giving Microsoft a unique AI cloud advantage. Although this approach makes sense from a business perspective, it raises serious issues with decreased substitutability and innovation stagnation from the perspective of the market ecosystem. By denying competitors the chance to improve their algorithms, a dominating digital enterprise can easily impede their progress by data collusion. Future developments might be directed to benefit a select group of gatekeepers rather than the general public if they must fit into Microsoft's ecosystem in order to be viable. Further, a single link failure (such as a Microsoft outage or API change) can cause an entire AI workflow chain across industries to cease functioning. The Commission also noted that data further strengthens and entrenches the network effects limiting inter platform competition.

In the Belaire Owners' Association vs DLF Ltd., CCI took “dependence of customers” as an important factor to attribute dominance. Given the limited interchangeability of LLM services between the various other providers, each user is being locked into a particular ecosystem & the particular firm continuously provides new innovation to keep the consumers hooked into it. 

For example - OpenAI’s integration into Microsoft 365 CoPilot is an example of ecosystem lock-in. Due to internal reliance on particular APIs and data flows, switching can be expensive for businesses once embedded in this ecosystem. In the same way, companies are completely locked in to Microsoft's environment because OpenAI's models are only restricted to Microsoft cloud only.

Although this approach makes sense from a business perspective, but it raises serious issues with decreased substitutability and innovation stagnation from the perspective of market ecosystem. Future developments might be directed to benefit a select group of gatekeepers rather than the general public if they must fit into Microsoft's ecosystem to be viable. Further, a single link failure (such as a Microsoft outage or API change) can cause an entire AI workflow chain across industries to cease functioning.

Conclusion

The Silicon Valley cliche that smaller players can always upend large ones is refuted by this. The “false comfort” in AI is the belief that the market is still open since the entry barrier is so high and the ecosystems are so closed. In actuality, we are quickly approaching a system of platform capitalism in AI, in which a small number of players control access to the user base, training data, and critical infrastructure.

Thus, such practises significantly distort market dynamics by limiting competition, innovation and consumers choices, creating conditions which will cause long-term AAEC in the market, which directly impacts cost, sensitive, smaller developers, depriving them of cost-effective options and increasing their dependency, which restricts innovation and discourage arrivals from developing superior alternatives. 

This results in fewer advancements in AI technologies, reducing the availability of diverse high-quality and affordable solutions for consumers. The reduction in choice will lead to a loss to consumers, thereby leading to harm to consumers, which goes against the legislative intent of the Competition Act, which seeks to ensure that consumer interest is safeguarded by fostering healthy competition in the market. 

There is a rising need for multidisciplinary review of regulators in data, telecom etc with the competition regulators with rising dimensions of such markets into multidiscipline. Many mature competition jurisprudence adapted ex-ante regulations to address these data related concerns & gatekeeper status of some online platforms. Fast changing markets require timely decisions to avoid this more abuse of dominance in such markets. Recently, CCI has launched market study into the AI market, which is a positive step towards this, but there is a need for CCI to scrutinise these firms taking defence under nascent nature of market.

About the Authors

Shivam Singh and Priyanshi Jain are 4th-year B.A. LL.B. (Hons.) Students at Dharmashastra National Law University, Jabalpur.