OPINION | Ensuring a Level Playing Field: How regulation can prevent AI from stifling competition

-

Aalok Thakkar

In my computer science course, I tasked students with using AI tools to code. The idea was to modernise and democratise learning by giving everyone access to AI assistance. Yet the outcome was the opposite: students with premium subscriptions outperformed those using free models. What was meant to level the field instead exposed a new digital hierarchy.

This mirrors a larger reality: as India's MSME and startup ecosystem gains momentum, AI promises to accelerate growth and innovation; the challenge is to ensure it doesn’t also accelerate the concentration of power.

The Competition Commission of India’s (CCI) Market Study on AI and Competition, released on October 6, addresses this moment directly.

AI-facilitated anti-competitive practices

The CCI study highlights algorithmic collusion as an immediate concern. AI-driven pricing algorithms can align market behaviour on their own without human intervention.

Within India, 37% of AI startups surveyed by the CCI already perceive AI-enabled tacit collusion as a market reality.

Data concentration poses a deeper, structural threat. Nearly 68% of Indian AI startups cite access to large, high-quality datasets as their biggest barrier. The imbalance is self-reinforcing: firms with more users collect more data, train better models, and attract even more users.

Infrastructure is also highly concentrated

Infrastructure is similarly concentrated: AWS, Microsoft Azure, and Google Cloud account for over 60% of India’s cloud market. The CCI’s findings reveal that 67% of Indian startups build applications atop existing foundation models, while only 3% work on developing foundation models themselves.

Risks arising from vertical integration

Self-preferencing through vertical integration compounds these risks. Dominant firms can use proprietary algorithms to prioritise their own products or services, steering users toward closed ecosystems. AI also enables precision predatory pricing, selectively undercutting competitors for specific consumer segments. Such micro-targeted strategies can erode competition before regulators even detect the pattern.

Regulatory interventions

India’s competition framework is adapting to the realities of algorithmic markets. The Competition (Amendment) Act, 2023, introduces two provisions of particular relevance: Section 3(3) formally recognises hub-and-spoke cartels, allowing liability to extend to firms that enable collusion indirectly (such as digital intermediaries), and Section 5 mandates that any merger or acquisition exceeding ₹2,000 crore be notified to the Competition Commission of India (CCI), regardless of turnover or asset value. These close a gap in digital markets, where firms acquire data-rich startups with negligible revenue and disproportionate strategic value.

The CCI’s study further recommends algorithmic transparency (requiring explainability for automated decisions), data portability, interoperability standards to prevent lock-in, and enhanced scrutiny of data-driven mergers. These measures align with India’s broader digital governance framework, including the Digital Personal Data Protection Act, 2023, which regulates the lawful processing of personal data, and MeitY’s AI Governance Guidelines (2025).

Taken together, these initiatives point to a regulatory philosophy still coalescing but directionally clear: data asymmetry is both a competition and a governance problem.

Government Initiatives: The IndiaAI Mission

The ₹10,300 crore IndiaAI Mission is India’s most ambitious attempt to address the structural gaps identified in the CCI study: limited compute capacity, restricted access to high-quality datasets, and an acute shortage of skilled talent.

It proposes shared GPU and TPU infrastructure for startups and MSMEs, curated public datasets, open-source foundation models, subsidised cloud credits, and AI literacy programmes to reduce dependence on a few global technology providers.

The mission’s success, however, will depend less on the allocations and more on the governance structure. Shared compute facilities will be useful only if access is transparent and efficient, and public datasets will be useful only if regularly updated, standardised, and open for reuse. It must also address the talent bottleneck that stems from weak research linkages between academia and industry, fragmented university funding, and limited incentives for interdisciplinary work.

The IndiaAI Mission thus sits at a crossroads. Its outcome will hinge on whether the initiative can move beyond distribution of resources to building transparent, research-oriented institutions capable of sustaining open and competitive AI development over the long term.

The path forward

India’s AI market has expanded from $3.2 billion in 2020 to $6.05 billion in 2024 and is projected to reach $31.9 billion by 2031. The window for preventive regulation is closing fast. Once market concentration solidifies through algorithmic lock-in and vertical integration, reversing it becomes exponentially harder. Unlike the EU or the US, we have the rare opportunity to shape competition policy in tandem with the growth of our AI ecosystem.

What happened in my classroom need not mirror India’s future. Avoiding that outcome will require the regulatory vigilance and infrastructural equity that the CCI’s analysis and the IndiaAI Mission have only begun to outline, and that future policy must now deliver in full.

 Aalok Thakkar is Assistant Professor, Ashoka University. Views are personal and do not represent the stand of this publication.