1. The Competition Commission of India (“CCI”) released its much-awaited market study on the Artificial Intelligence (“AI”) sector in India on 6 October 2025.

2. The market study describes the various technological layers involved in the AI ecosystem, called the ‘AI stack’. The AI stack is organized as layers that represent the entire AI workflow – “upstream” and “downstream”.  The upstream AI stack (where the data and foundational technologies are prepared) involves data, infrastructure, development and foundation model layers. The downstream AI stack (where AI gets adapted and deployed in real-world contexts) involves the AI model, release and deployment, user interaction and governance & orchestration layers.

3.The market study observes that the Indian startups are largely present in the AI model layer but not in the upstream AI stack which is dominated by large multinational technology companies such as Alphabet, Amazon, Meta, Microsoft, Open AI and NVIDIA.

4.The key findings and recommendations in the market study are below:

Potential competition concerns with AI

5. Algorithmic Collusion: The use of AI algorithms may lead to tacit collusion amongst enterprises as self-learning algorithms can independently adopt cooperative pricing strategies that maximise profits. Algorithms can also collude by allocating markets, limiting output, or even coordinating on bidding strategies in online markets.

6. Algorithmic Unilateral Conduct: Dominant enterprises may deploy AI algorithms to engage in exclusionary and exploitative conduct –

1. Self-preferencing: Search ranking algorithms may facilitate self-preferencing by favouring own related entities. Separately, dominant entities may indulge in self-preferencing of own products across the vertical AI tech stack.
2. Predatory pricing and price discrimination: Since AI algorithms can monitor pricing in real time, predatory pricing or price discrimination strategies may be implemented successfully by targeting customers who are price-sensitive or prone to switching.
3. Tying and Bundling: Similarly, AI algorithms may be used to identify price sensitive customers to offer discounted bundles. Further, Big Tech enterprises may bundle their AI applications with their core products such as search engines, browsers, operating systems etc.

7. Pricing practices: AI algorithms have the ability to provide dynamic, targeted and personalized prices based on data inferred from consumer preferences, brand loyalty, purchasing behaviors etc. A dominant enterprise may target its rivals’ customers and nudge them with selective, lower, and targeted prices. This may also lead to loss of trust from customers in the online markets; thereby increasing search and transaction costs.

8. Market structure issues – entry barriers, network effects and reduced transparency: Significant entry barriers were identified such as availability of data, cost of AI infrastructure (e.g. GPUs and cloud services) and availability of skilled resources. It was also highlighted that the incumbents in the upstream AI stack would benefit from indirect network effects as downstream players, new entrants and end-users are likely to be dependent on their applications and tools. With opaque and unclear processes of the AI algorithms, there may be increased dependency on incumbents as well as uncertainty. 

Recommendations

9.The market study recommends the following measures to develop competition compliance, promote innovation and ensure fair competition across the AI stack:

1. Self-audits: Enterprises are urged to conduct self-audits of AI systems for competition compliance. A guidance note is also provided for enterprises to create their own tools for self-compliance.
2. Increased transparency: Enterprises must reduce information asymmetry by communicating about: (a) the usage and purpose of AI decision making; (b) the main parameters behind those decisions; and (c) any other information that fosters trust in AI-powered systems. This does not however, require enterprises to disclose their algorithms or proprietary / confidential information.
3. Removing entry barriers: Stakeholders are urged to address the entry barriers by expanding national AI computing infrastructure, promote open-source AI frameworks, develop data repositories, equip India’s workforce with technical expertise etc.
4. Focused advocacy and capacity building: The CCI intends to undertake advocacy efforts by conducting a conference on “AI and Regulatory Issues” and focused workshops on competition compliance. It also intends to set up a think tank consisting of domain experts to assist the CCI.
5. Inter-regulatory and international cooperation: The market study recognizes the need for coordination amongst the CCI and authorities dealing with data protection, cybersecurity, and intellectual property laws. It also emphasized international cooperation with other competition authorities and multilateral platforms such as the OECD, ICN and UNCTAD.

Conclusion 

10. The AI sector in India (and indeed across the world) is at a very nascent stage of development. In its market study, the CCI has employed a ‘light touch’ approach by identifying the potential issues that may arise in the sector while at the same time providing a road map with guidance for self-compliance. The CCI should be credited for accepting the need for it to develop its own technical know-how and expertise in this rapidly growing sector. The market study therefore demonstrates the CCI’s calibrated and non-interventionist outlook towards emerging technology markets, in line with the objectives of the Indian government.

11. Notwithstanding the same, the market study also identified several recent acquisitions in the AI sector in India and highlighted the need to review ‘killer acquisitions’ and partnerships in the market. The CCI can therefore be expected to continue with its rigorous scrutiny of M&A transactions, aided by the recently introduced deal value thresholds. 

Authors: Sonam Mathur – Partner; Shubhang Joshi – Managing Associate and Saikishan Rathore – Associate 

Disclaimer: This alert only highlights key issues and is not intended to be comprehensive. The contents of this alert do not constitute any opinion or determination on, or certification in respect of, the application of Indian law by Talwar Thakore & Associates (“TT&A”). No part of this alert should be considered an advertisement or solicitation of TT&A’s professional services.