GenAI calls for a relook at what constitutes anti-competitive practice

The CCI’s market study on AI and competition must focus on the competition issues and leave the ethical and data dimensions to be addressed by other regulators and departments of the government.

By Samir R Gandhi and Binit Aggarwal 

There have been remarkable advancements in the development of artificial intelligence (AI) in the last few years. Foundation models, which are trained on a broad set of unlabelled data that can be used for different tasks, with minimal fine-tuning like OpenAI's ChatGPT, have become the technological base for creators and entrepreneurs to develop customised and targeted AI solutions and have attracted significant interest and investments.

Predictably, the widespread use of AI, has also presented regulators with questions surrounding the ethical use of AI and affixing accountability for its potential misuse. In this context, competition regulators worldwide have started to examine whether the proliferation of AI foundation models could distort competitive conditions in the marketplace and the Indian market regulator- the Competition Commission of India (CCI), is no exception.

Common quest of regulators

The CCI has recently commissioned a market study on competition in the AI market to better understand its dynamics, on the heels of another market study commissioned by the UK’s Competition and Markets Authority (CMA) on AI Foundation Models.

The CMA first published its report in September 2023, subsequently updated in April 2024, where it recommended targeted intervention in order to prevent overregulation. The CMA report discusses in particular, three notable aspects of AI foundation models, including:

Building models are expensive and resource intensive

The CMA identifies the considerable investments and data required to build AI foundation models and points out that invariably, it is companies which have access to significant resources and that are best placed to access large data sets and make such investments.

It provides first-mover advantage

This resource advantage could allow existing players to extend their positions of market leadership into AI markets and disadvantage competitors and developers who don't have access to such resources. The resource advantage could allow such companies to integrate their existing services with proprietary AI foundation models, and restrict consumers from choosing their preferred AI providers.

AI creators enjoy advantages in other tech markets too

The CMA also highlights that the companies who own and control AI foundation models could misuse their position to consolidate their strength in other markets that depend on AI enabled services, such as retail, e-commerce, logistics and delivery, etc. This would disadvantage smaller firms, who may lack the AI fire-power and are unable to compete effectively in markets which depend on AI enabled services.

A web of contracts can achieve consolidation without mergers

Competition regulators review M&A deals to ensure that markets remain competitive, but the CMA has pointed out that partnerships such as the one between Microsoft-OpenAI, could be used by large technology companies to consolidate their presence in the AI market and may not be reviewed under traditional competition M&A rules. For instance, the European Commission, which had earlier cleared Microsoft’s OpenAI investment, is now having a closer look at the deal to ensure that partnerships do not allow one partner to control or influence the other.

Pointers for the CCI’s market study

The CCI should be complimented for having commissioned a study to understand what (if any) competition regulation is required to keep markets competitive with the proliferation of AI. The CMA study takes note of the delicate balancing act between promoting AI driven innovation and regulation, and suggests a set of AI principles, which include open access, diversity of models, ability to choose, fair dealing, transparency, and accountability.

The CCI through its own market study will undoubtedly understand and study the risks and rewards of deploying its own competition tools to address the potential misuse of AI foundation models in India’s vibrant tech market.

The CCI has a range of “fit for purpose” tools that it utilises to examine anti-competitive behaviour under the Competition Act, which have been further augmented through recent amendments.

The Commission, has in the past, used market studies to inform itself of market realities, engage with stakeholders and suggest “self-regulatory” measures for market participants where appropriate. Such market studies enable the CCI to be effective and agile, should anti-competitive behaviour be brought to its attention or detected in the market.

Much like its global counterparts, the CCI’s market study on AI and competition must focus on the competition issues arising out of the proliferation of AI foundation models and leave the ethical and data dimensions presented by this complex subject to be addressed by other regulators and departments of the government. Further, the CCI’s study would be greatly strengthened if it were based on the specific needs, discussion and context of the wider Indian tech ecosystem and its billion consumers.

Samir Gandhi is Partner & Co-Founder at Axiom5 and Binit Aggarwal is associate at Axiom5. 

Views are personal and do not represent the stand of this publication.