Algorithmic collusion and the limits of the Indian Competition Act

Author- Srishti Singh

Abstract

The rapid inclusion of the large language models and particularly the automated decision-making system, has now changed how the big corporate firms will determine and regulate the prices in the market. Nowadays, the firms use the pricing algorithms, which are trained to monitor the market and adjust the prices in real time, without any type of human interaction for the same. While these practices amount to unfair trade practices and ultimatelykeep the customers at a disadvantage, and grant the firms huge profits.  This blog argues that the Indian act to regulate unfair trade practice in the market, that is, the Competition Act, 2002, lacks the competence to address the changes emerging with the growing inclusion of Artificial Intelligence in the market.

Introduction

In India, we have the Indian Competition Act, 2002, simply built on the assumption that all breaches of fair-trade practices result from human coordination. These included activities such as, price- fixing, market allocation, limiting output, etc. Earlier, Cartels required explicit agreementsor mutually agreed agreements that required communication and conscious strategic actions between the various players of the markets. For example, two leading firms mutually agreed that they will not keep their price below a certain amount, and as a result, consumers are forced to pay a high price for that particular commodity.

To tackle this, we have Section 3(3) of the Competition Act, 2002, that prohibits agreements made by businesses that are in the same or similar lines of business, that such agreements would result in an adverse effect on the competition and ultimately the buyer is at a loss because they would be forced to pay a high price. Both in India and in the world, the practice of cartel is punishable or penalised, when the essentials of an established communication, either direct or circumstantial evidence is present.

Understanding Algorithmic Collusion

However, now in the 21^st century, with the rapid amalgamation of Artificial Intelligence in the digital market, it challenges the foundational assumption of cartels that they need communication for the same.  Now, these price fixings are not done by individuals through the agreements, but rather are done by the large language models as they are highly capable of learning, adapting and responding automatically to the market conditions. Industries, such as e-commerce, ride-hailing services, airlines, and hostel services, use price-fixing algorithms. These algorithms are trained to monitor current market conditions, keep an eye on the prices of their competitors and set the prices accordingly in the real time.

It is very evident that this algorithmic collusion does not require any prior agreements, and hence, the firms nevercommunicate or mutually agree to fix prices. As a result, these markets will show cartel effects like higher advance prices andless dispersion in pricing among competitors. This collusion by Artificial Intelligence models poses a challenge in the interpretation of Section 3 of the Competition Act.  At the time of determining whether the mutual coordinationcould amount to unfair trade practice under section 3 (3) of the Competition Act, 2003.

Over time, three forms of algorithmic collusion have been identified. First, monitoring algorithms- with the help of this algorithm, firms can detect and respond to the changes in price in no time. This enables them to respond intensively to any change in the pricefluctuation. Second, is the parallel algorithm- when more than one firm depends on the same software to generate pricing outcomes. Third, and most used algorithm is the self-regulatingalgorithm. It is an autonomous model and decides its prices, relying on the repeated market interaction.

By using these AI tools, the same purpose is served by the firms, only the difference is that it is hard to make them liable under the traditional anti-competition laws.

Challenges to the traditional Competition Law

Section 3 (3) of the Competition Act prohibits and penalises such agreements that lead to the fixed price fixing and keep customers at a disadvantage, as they are forced to pay high prices. But now, as the time has evolved, it is very difficult to make the firms liable for algorithmic collusion. First, the essentials of an existing agreement could not be proved because Algorithms do not form any agreement among themselves.

Second, it is very difficult to prove the firms ever had any intention to make such agreements, as traditional cartels depend on evidence such as emails, letters, etc., but now in the present times, no such evidence exists, and hence the firms could not be made liable under the said section of the Competition Act.

Third, and the most important, is who will be made liable for any such anti- competitive acts. Whether the company developing the system is held liable, or the developers whose idea it was, or the firm that used that software to engage in the anti-competitiveness act, shall be held liable.

Way Forward

India is rapidly transforming towards the digital economy and in the age of Artificial Intelligence, and hence, this algorithm-driven coordination has become extremely relevant and needs immediate attention. Though the Competition Commission of India (CCI) has took proactive role and tried to examine the issues on its part, but the statute, i.e., the Competition Act, 2002, still focuses on the traditional approach for tackling the algorithmic collusion. With the passage of time, India has shown a strong growth in data-driven industries, and relying only on traditional statutes may prove insufficient.

The Competition Act should undergo several changes or amendments to bridge the gap between the algorithmic collusion and the traditional statute. As one of the primary defense to escape the liability from competition act is the lack on intention on part of the firms, ass the prices are decided by the AI models, however, Algorithms are the integral part of the firm’s structure and they are always set- up and executed by the firms itself and hence, should not be exempted from the liability, they would have faced when would have indulged by a anti- competitive act by forming proper agreements.

Another solution would be that the parliament would make amendments and include clear and explicit essentials to make the organisers liable if they get involved in algorithmic collusion or any other practice related to anti-competitiveness.

Conclusion

Algorithmic collusion has emerged as a fundamental challenge that needs to be tackled in order to promote fair competition in India in this digital age.  As these large language models tend to influence market behaviour,and that without direct human interaction so far.The existing legal framework and statutes designed to regulate cartelsare not sufficient and have some shortcomings.  And hence the evolution of digital markets, hence demands an evolution and rectification in competition regulation.

The competition law must adapt to the evolving needs of the society and need amendments to include the autonomous systems capable of reshaping the competitive dynamics and protecting the people at the receiving ends, that is, the customers, from exploitation by the hands of the big businesses gaints.

References

1. Arpita Gupta, Algorithmic Collusion and Its Challenges to Antitrust Regulations, International Journal for Legal Research and Analysis, available at: https://www.ijlra.com/details/algorithmic-collusion-and-its-challenges-to-antitrust-regulations-by-arpita-gupta⁠� (last visited Mar. 5, 2026).

2. Centre for Business and Commercial Laws, available at: https://share.google/9eOL8RhvoD2OARKL� (last visited Mar. 5, 2026).

Note - The information contained in this blog is for general information purposes only. We endeavour to keep all the information up to date and try our level best to avoid any misinformation or any kind of objectionable content. If you found any misinformation or objectionable content on this website, please report it to us at editors.ilw@gmail.com