CCI Market Study on Artificial Intelligence and Competition (2025)

The CCI began preparing for a study of AI markets in 2024. In April 2024 it issued a request for proposals for a market study on AI and competition, seeking to understand emerging competition concerns in the development and deployment of AI systems and their effect on innovation in key sectors. The RFP listed topics such as market concentration, entry barriers, AI-enabled collusion, algorithmic bias, AI-driven pricing and recommendations, and mergers involving AI-capable firms. Proposals were due on 3 June 2024, and a public notice of clarifications on the RFP followed.⁵⁶⁷ The study was carried out for the Commission by the Management Development Institute (MDI), Gurgaon, which gathered the data through surveys, stakeholder interviews and focus groups.¹

In August 2025 MediaNama reported that the study was at an advanced stage, citing the Parliamentary Standing Committee on Finance's report on the CCI's role in the digital economy. The committee said the CCI was being equipped to handle AI risks such as algorithmic collusion and data monopolisation, that the study should serve as foundational evidence for refining the draft Digital Competition Bill, and it asked about plans to regulate agentic AI.⁸ The report was published in early October 2025, and MediaNama's first coverage is dated 7 October 2025.³

The report describes itself as an exercise under the Commission's "broader advocacy mandate". It carries a disclaimer that it is not a legal document, may not be used for any proceedings and is based on broad trends and views that emerged in the study.¹

The study cites market-size figures showing the worldwide AI market growing from USD 93.24 billion in 2020 to USD 186.43 billion in 2024 and India's from USD 3.20 billion to USD 6.05 billion, with India expected to reach USD 31.94 billion by 2031.¹

It describes the ecosystem as a stack. Upstream are the data layer (collection, labelling, management), the infrastructure layer (servers and specialised chips through cloud and edge computing), the development layer (algorithms, tools, foundation models) and the generative-AI and model layers. Downstream are fine-tuning and deployment for specific tasks. The study notes that global firms such as AWS, Microsoft Azure and Google Cloud are the key cloud providers, that NVIDIA, Intel and AMD are the strongest players in high-end AI chips, and that OpenAI, Microsoft, Google and Meta are the major foundation-model players, while Indian startups are making inroads in various layers.¹

Primary research found that around 67% of the startup respondents work mainly on building AI applications, that 76% build on open-source technologies, and that 88% use machine learning, 78% natural language processing and 66% generative AI. Among user industries, AI is used mostly for customer-facing functions, with about 90% of respondents using it to monitor customer behaviour and 69% for demand forecasting. Sector adoption was highest in banking, financial services and insurance, healthcare, IT and manufacturing (52%).¹

The study also records pro-competitive effects. AI helps adopters reduce costs and improve efficiency, non-adopters may struggle to match the personalisation and responsiveness of AI-driven rivals, and the report has a section on pro-competitive aspects for MSMEs.¹

The report groups the main concerns as follows:¹²

On collusion, the study classifies pricing algorithms as monitoring algorithms that collect competitor data, parallel or hub-and-spoke algorithms where several firms use a common vendor, signalling algorithms that respond to rivals' moves, and self-learning algorithms that use reinforcement learning to maximise profit. It observes that the last two "may not have been designed to achieve collusion" but can reach collusive outcomes depending on market conditions, and that stakeholders disagreed about whether similar prices necessarily signal collusion.¹⁴

The startup survey indicated that 37% of respondents saw a possibility of AI-facilitated collusion, 32% of price discrimination and 22% of predatory pricing.¹

The study frames competition law as sector and technology agnostic and able to address many AI-driven practices. It points to the Competition (Amendment) Act, 2023, which introduced provisions on hub-and-spoke cartels, deal-value thresholds and settlements and commitments, as equipping the CCI for such challenges, and to the Digital Personal Data Protection Act, 2023 as a complementary measure.¹

The action plan in Chapter 6 combines measures for firms and for the Commission:¹²

1. Self-audit of AI systems. Enterprises, particularly those with market power or wide consumer reach, are urged to adopt self-audits. The indicative principles include documenting AI-based decision-making (objectives, data sources and access protocols), designing and testing algorithms with safeguards against unintended anti-competitive outcomes, regular internal algorithmic audits, review of AI-driven pricing for unintended price alignment or discrimination, safeguards when using third-party tools, and measures to prevent sharing commercially sensitive data among competitors. Annexure I provides a guidance note, which MediaNama describes as a six-step process from forming a cross-functional team and mapping AI systems to ranking risks, auditing, recording findings and re-evaluating.¹³
2. Transparency. Firms are encouraged to communicate the use and purpose of AI in decision-making and a general description of the main parameters, in plain language and updated periodically, without disclosing proprietary algorithms.¹
3. Focused advocacy. A conference on AI and regulatory issues and workshops on AI and competition compliance.¹
4. Removing entry barriers. Continued government measures to expand access to AI infrastructure and capabilities, including compute, open-source frameworks, datasets and skilled workforce.¹²
5. Regulatory capacity. Strengthening the CCI's technical capabilities and setting up a think tank on digital markets with a focus on AI.¹
6. Inter-regulatory coordination. Cooperation with other regulators; MediaNama notes this includes MeitY and the Data Protection Board through memoranda of understanding under the Competition (Amendment) Act, 2023.²
7. International cooperation with foreign competition authorities and forums such as the OECD, UNCTAD and the International Competition Network.²

The report adds that any allegation of anti-competitive conduct will still be examined by the Commission case by case under the Competition Act, 2002.¹

MediaNama's analysis pointed out that most of the measures are advisory rather than binding, that self-audits and transparency rely on voluntary participation with no enforcement mechanism or reporting deadlines, and that several recommendations, such as infrastructure and skills, depend on ministries other than the CCI.² It also noted that the CCI had not said how it would monitor voluntary audits or verify internal reviews, and contrasted the approach with the EU's conformity assessments for high-risk AI systems, describing the CCI proposal as softer and focused on internal accountability.³

The study is also expected to feed into competition policy more broadly. MediaNama reported that the Ministry of Corporate Affairs said the study would help set enforcement priorities under the proposed Digital Competition Bill, and that the study is one of several pieces of evidence the Ministry cites alongside CCI studies on e-commerce, telecom and other sectors.³⁹

As of 15 October 2026 the market study remains the CCI's principal published analysis of AI and competition, and the self-audit and transparency measures remain voluntary. The sources reviewed for this entry did not identify any CCI enforcement action or regulation adopted specifically on the strength of the study.

The Digital Competition Bill, which would impose ex-ante obligations on large digital enterprises, was still pending in August 2026. MediaNama reported on 18 August 2026 that the Government had not finalised the draft and was awaiting a separate market study commissioned by the Ministry of Corporate Affairs on thresholds for big tech companies and core digital services, for which a request for proposals was floated on 3 November 2025 with a bid deadline of 5 January 2026. The same report notes that the Government cited the CCI's AI study as part of its evidence base.⁹ The Parliamentary Standing Committee on Finance had earlier observed that AI agents might fall outside the draft Bill's list of core digital services, and recommended including them.⁸