Rethinking The Idea Of ‘Agreement’ In Section 3 Of The Competition Act For An Age Of Pricing Algorithms
Ayush Agrawal, Adarsh Sahu & Ankul Prajapati, National Law Institute University, Bhopal
ABSTRACT
This article asks one narrow question. Has the concept of ‘agreement’ that has anchored Section 3 of the Competition Act, 2002 for two decades built on proof of a communicated meeting of minds between economically independent actors remained adequate now that firms can converge on supra- competitive prices through pricing algorithms and platform architecture, without any human ever communicating an intention to collude? The answer defended here is that it has not. The Competition (Amendment) Act, 2023’s codification of hub-and-spoke liability is a necessary and well-designed response to the evidentiary gap exposed in Samir Agarwal v Competition Commission of India, but it resolves only one of at least four analytically distinct forms of algorithmically mediated coordination identified in the comparative literature on this subject. It leaves the hardest case autonomous convergence produced by self-learning pricing systems, absent identifiable human facilitation or intent entirely outside the reach of a doctrine still organised around the discovery of a communicated understanding. Building on Ariel Ezrachi and Maurice Stucke’s four-model taxonomy of algorithmic coordination, on the reasoning of the Court of Justice of the European Union in Eturas, on the first American criminal prosecution for algorithmic price- fixing, and on India’s own 2023 reforms and its still-pending Digital Competition Bill, this article argues that the correct legislative response lies neither in wholesale transplantation of the European Union’s ex ante gatekeeper model nor in continued reliance on ad hoc adjudication under an unmodified Section 3, but in a calibrated, rebuttable structural presumption aimed specifically at algorithmically mediated parallel conduct, joined to the institutional capacity to detect and audit it.
