Pricing Algorithm and Artificial Intelligence: A Competition Law Conundrum


The dynamic atmosphere of antitrust law has been witnessing effects of changes that are taking place in the arena of artificial intelligence. One such phenomenon has been algorithms and their usage by companies to further their profit maximisation. These algorithms are continuously utilized by e-commerce companies to perform series of tasks automatically with the help of the data that is fed to it. One such mechanism that plays an important role is the pricing algorithm. Pricing algorithms allow for constant changes in prices of products according to the area they serve and the data they collect. Over a short period of time, there has been a significant increase in the usage of these algorithms probably because of their effectiveness. Today a large chunk of businesses all over the world depend on these algorithms to subsist. The concern that rises from the use of these algorithms is regarding their self learning nature and the eventual threat they create of collusion. The gap in law, especially in India with respect to the role of artificial intelligence in competitive practices forms a large part of the debate on this issue. Whether the pricing nature of these algorithms create collusion or not and if so then how do we tackle something of which no tangible record exists, are the few questions that have been delved into in the subsequent writing.

Mystery associated with Pricing Algorithm & Artificial Intelligence

With the ever changing technological environment and the ever developing self learning algorithms, new challenges are being continuously faced by antitrust laws. This would not have been considered as a problem in competition scenario a few years ago. Now, with every single movement of a person being tracked with activities such as Internet of Things (IoTs), it has become increasingly easy for companies to gather and use data on people for their benefit.

Competition law prohibits any action that may lead to anticompetitive practice or reduction of the competition, may it be agreements regarding anticompetitive practice, abuse of dominance or mergers that hamper competition. The difficulty that lies with AI is proving the existence of such illicit agreements, where any enterprise would adopt the pricing algorithm of their competitor after being made aware of any developments. Such agreements would be anticompetitive in nature. Such agreements do not take place in a format that would be easy to trace, rather they are in the nature of automated algorithms, where it’s near impossible to prove collusion. Conscious behavior by the firms to ensure that there is equilibrium in the pricing does not come under flak from the existing provisions. Here, the job of the antitrust authorities becomes to ensure that such algorithms and their developers must come under the notice of the government that consciously give platform for tacit collusion to happen.

The issue is that the competition agencies do not have appropriate tools to do so. These practices can come under the umbrella of unfair trade practices. ‘Anticompetitive intent’ is a presumably strong ground to establish behavior similar to a cartel, and therefore there needs to be legislation that can not only effectively address the transparency but also counter any abuse of such transparency by the competitors. Although more questions arise in such a situation with respect to legal tacit collusion and how will it be looked at when others are believed to violate antitrust laws. The probable solution here is to hold the online companies liable if they were motivated to act towards any activity that would result in chilling effect and was anticompetitive in nature or they were fully aware about the anticompetitive consequences of the pricing algorithms and still effectively participated.

The question that consequently comes up is, whether it’s possible to design an algorithm that operates with the appropriate checks and balances while at the same time ensuring that the profit maximization motive is not defeated? The answer is to this is murky given the massive data that these algorithms exists on. Another approach here could be to entrust the regulatory authority to ensure that they get regular information on the algorithms being used in the market, this would help determine the transparency they create in the market. However it would be visible only through practice on how the enforcement authorities and regulatory bodies will react to such futuristic challenges posed by AI in antitrust.

Competition Commission of India on Pricing Algorithm & Artificial Intelligence

In Samir Agrawal v ANI Technologies Pvt. Ltd., the Commission had the opportunity to decide on the usage of pricing algorithm in this case. Here the question was with respect to whether the utilization of a pricing algorithm by drivers using a common cab service platform amounts to cartelization under the act.

One of the submissions made by the Informant against the drivers connected to Ola/Uber’s network had been that they were not technically employees but rather third party service providers working independently. According to this structure, it can be concluded that the network works as a ‘Hub’ whereas the drivers work as ‘Spokes’, who eventually colluded on prices. They also submitted that the drivers do not function as a single economic entity with the service providers seeing as how they are independent of an employee/employer relationship. The Informant therein stated that the collusion by the drivers, which is inadvertently supported by the service providers (Ola/Uber), falls under ‘concerted action’ as provided under Section 3(3)(a) read with Section 3(1) of the Act.

The hub and spoke agreement is understood as exchanging of critical information regarding data and prices between competitors; this is facilitated through a third party and ends up showcasing a cartelistic conduct of the said competitors. The pre requisite for the arrangement to qualify as hub and spoke was the involvement of a third party who would help the spoke in creating necessary measures to not only to share sensitive information regarding the prices but also to fix predetermined prices. Thus it’s necessary for collusion or conspiracy to exist for there to be fixing of prices, and this conspiracy or collusion becomes imperative to prove the existence of hub and spoke arrangement.

The Commission eventually held that the prices that are determined by such platforms through the usage of algorithms would not be said to be collusion between drivers. They stated that for the hub and spoke arrangement to exist all of the drivers would be required to coordinate as a cartel to ensure that the prices are set accordingly. The Commission eventually stated that since no agreement of such sort existed where the pricing was delegated to the respective platforms, no violation under Section 3(3)(a) of the Act exists. They further stated that the prices were fixed by the algorithms in the App on the basis of large amount of data and several other situational factors. They held that the prices determined for different consumers varied because of the usage of algorithm and would not fall under the conventional understanding of hub and spoke arrangement. However, this explanation of usage of algorithm for collusion may not be the most appropriate.

Pricing Algorithm & Artificial Intelligence: The Road Ahead

The question that needs to be now addressed is whether the competition is being promoted by the increasing usage of pricing algorithms. The suggestion provided often for this is a necessary shift of technology towards regulatory and intervening mechanism. An example of this is the price mechanism of Uber wherein the algorithm decides the base price according to the distance and area as well as when to create a hike in pricing, for how long and to what extent. Defense claimed by Uber is that the price is determined according to the market demand and supply dynamics, the solution suggested for the same is that governments should also utilize these pricing algorithms to ensure an effective market price is set, using the same Big Data and analytics that companies depend on. This will not only ensure an equal stage for competition but also increase the reliability of the government through the algorithms they use.

The question that rises here is whether the current legal position regarding antitrust be applicable in virtual competition? Despite there being noticeable harm, it would be difficult to ascertain the violation in many cases. An example here is the Hub and Spoke issue explained above, which could be considered as an anti-competitive practice by bringing it under Article 101 of the Treaty on the Functioning of European Union (TFEU), Section 3 of the Competition Act, 2002 (India) and under Section 1 of the Sherman Act (US). However, Artificial Intelligence makes it increasingly difficult to identify contributors of collusion or any other anti-competitive practice. The geographical market and relevant product being the major determinants on which the pricing algorithm survives also makes it relatively difficult to pinpoint where the clear market power lies. Virtual competition should be made comprehensive enough that it promotes healthy competition and doesn’t infringe upon the privacy of an individual while also extending consumer welfare.

Hence the proposal lies towards creation of such an legislations that lets people have control over their data and also provides them right to know when they are being tracked by these networks, offline or online. The EU Commissioners’ speech in 2016 becomes essential here wherein he stated that competition law may not be the solution to all the problems but it can provide important contributions towards keeping digital markets fair and equal.


  1. Mark L. Krotoski, Key International Tools Used to Investigate Cartels and Enforce the Sherman Act Abroad, Int’l Antitrust Bull., at 18-21 (Mar. 2015).

  2. Daniel Mandrescu, ‘Applying EU Competition law to online platforms: The Road Ahead-Part I’, European Competition Law Review, 2017, 38 (8), 353-365, at p. 4

  3. Julian Nowag, ‘The UBER-Cartel; UBER between Labour and Competition Law’ Working Paper Series: LundLawCompWP 1/2016.

  4. Ezrachi, A and M E Stucke, “Artificial Intelligence and Collusion: When Computers Inhibit Competition”, Oxford Legal Studies Research Paper No. 18/2015, University of Tennessee Legal Studies Research Paper No. 267 (2015).

44 views0 comments

Recent Posts

See All

Recently Apple’s years-long antitrust dispute in South Korea came to an end when the Korean Fair Trade Commission (KFTC) accepted Apple’s proposal of the a financial package of US$89.73 million in ful