Collusion, gatekeeping in AI: CCI calls for industry self-regulation, self-audits

According to the study, India has a number of regulatory remedies to deal with the rise in use of AI. However, the report emphasised a clear need for proportionate safeguards, transparency standards, and industry-led self-regulation to ensure the safe, ethical, and inclusive use of AI technologies. To encourage fair competition, the CCI has also recommended that enterprises should carry out self-audits by documenting their governance processes.

One of the foremost issues in application of AI is the risk of algorithmic collusion, where enterprises use AI-driven pricing algorithms that may learn to align prices over time. This can lead to outcomes similar to price-fixing, even in the absence of direct human coordination, making detection and enforcement by regulators far more complex, the study said.

In a survey conducted while framing the study, AI-facilitated collusion was the biggest concern across stakeholders, followed by price discrimination and increased entry barriers. The concerns are compounded as over one-third (67 per cent) AI start-ups in India are focused on the application layer, rather than the core foundational layer that powers them.

Big Tech advantage

“In the AI industry, major firms may leverage their control over data, infrastructure, and proprietary models to entrench their market position which may end up raising barriers to entry,” the study said. And the structure has so far evolved in a way that “positions big technology firms with a significant advantage in data access”.

It said that over years of operating large-scale platforms and services, big companies have accumulated vast datasets. Since new AI firms and start-ups increasingly rely on the infrastructure and cloud offerings of these hyperscalers, data often flows back to the incumbents (depending on the nature of the agreements). This reinforces their data advantage and skews data access in favour of big technology firms.

Enterprises that control vast datasets may be able to train more effective AI models, giving them a substantial competitive edge. This may entrench market power and create high barriers to entry for smaller or newer players, stifling innovation and reducing consumer choice, it added.

Self audits, documentation the way forward

The study — prepared by the Gurugram-based Management Development Institute (MDI) on CCI’s behalf — has proposed the introduction of enterprise self-audits to ensure fair competition across the AI development value chain and applications.

The audits would include documentation of AI-based decision-making process including algorithmic objectives, data sources, and access protocols. They would also include design and testing of algorithms with built-in safeguards to prevent/detect unintended anti-competitive outcomes/practices and reduce unfair competition risk. Regular internal algorithmic audits/periodic review of algorithmic outputs/market outcomes to proactively identify and eliminate inadvertent algorithm-driven collusion will also be included.

Noting that bigger enterprises gatekeeping how their algorithms work could cause market distortion in their favour, the study urged companies to adopt transparency measures to reduce information asymmetry. This includes documenting the usage and purpose of deployment of AI in decision-making, and a general description of the main parameters for AI-based decisions, among other things.

From the homepage

Bihar Assembly Election 2025 Live Updates: NDA alliance finalises Bihar seat sharing as BJP, JD(U) keep 101 seats each

‘Technical issue, no other problem’: Afghan Minister Muttaqi plays down exclusion of women journos from press meet

‘How did they come out at 12.30 am?’: Mamata Banerjee questions private medical college’s accountability after MBBS student’s rape