Vivek Agarwal
The Competition Commission of India (CCI) has published its Market Study on Artificial Intelligence and Competition, which has observed that AI algorithms can collude without human intervention and be used to engage in exploitative conduct.
Multinational tech companies – such as Alphabet, Amazon, Meta, Microsoft, Open AI and Nvidia – are dominant players in processing data, providing computer resources and creating foundational models. But startups face major barriers when competing with these firms. Competition law expert Vivek Agarwal, who heads the antitrust boutique firm CompLaw, discusses the CCI report. Agarwal helped draft CCI regulations and amendments to the Indian competition statute. He has been involved in key cases relating to cartels, abuse of dominance, vertical restraints and combinations.
Q1. What are your views on the report and on the CCI’s approach to AI?
The CCI’s market study is a wise initiative to understand competition law issues associated with the growing adoption of AI. It takes a balanced view on potential AI-related concerns and suggests self-regulation instead of hard regulation.
India is steadily emerging as a key player in the AI race. At this juncture, the CCI is right in not putting any premature regulation in place, which could decelerate innovation and impact India’s AI potential. The CCI followed a similar approach with its 2020 e-commerce market study report, which served as an important foundation for enforcement.
The AI market study contains critical insights for those who may be disadvantaged due to bigtech’s misuse of AI and may prompt them to raise their concerns before the CCI. The study could therefore pave the way for developing a realistic and calibrated enforcement regime in the AI space. The AI report is timely and the CCI could use it to engage with the government to ensure that the “under-construction” digital competition law is better equipped to effectively deal with potential competition issues arising from the use of AI.
Q2. What are the major challenges the competition regulator faces with AI?
Conventional competition rules may not be enough to catch contraventions like algorithmic collusion, which does not involve any human co-ordination. Self-learning algorithms or generative AI would further undermine traditional antitrust enforcement.
Another challenge would be dealing with high entry barriers for startups. It is difficult for them to access quality data, cloud services, skilled talent and funding. Bigtech firms have control over essential infrastructure, proprietary foundation models and vast proprietary data.
The CCI also needs to strengthen its institutional capacity and develop technical capabilities to maintain effective regulatory oversight of AI.
Q3. The CCI has asked companies to self-regulate on competition compliance in the AI domain. What do you think of this approach?
The CCI’s approach is pragmatic and aligned with global best practices on algorithmic accountability. The CCI rightly expects firms to embed fundamental competition law checks into their AI models. If these models are allowed to operate as ‘black boxes’, detection and prevention of AI-related antitrust issues may not be possible through ex-post enforcement.
Overall, the self-regulation framework is a reasonable starting point to monitor competition law risks in a rapidly evolving AI space.
Companies are deploying sector-specific AI solutions tailored to address their unique operational requirements. E.g., E-commerce companies deploy AI on demand forecasting, inventory optimisation and personalised customer experiences. Such solutions increase operational efficiency, allow data-driven decision making, improve service quality and may generate sustainable competitive advantages for early adopters.
Companies may also use AI-based pricing tools, which raises concerns over algorithmic collusion. Additionally, dominant enterprises may use AI tools to indulge in unilateral anti-competitive practices, such as, self-preferencing, predatory pricing and price discrimination.
Q4. What actions do you foresee the competition regulator taking? The CCI has observed that the AI industry offers incumbents a substantial and often insurmountable competitive edge and puts up high entry barriers for newer players.
The CCI could step up its advocacy initiatives by organising focussed workshops to sensitise stakeholders on competition compliance and foster pro-competition innovation. It could also sign MoUs [memorandums of understanding] with other statutory authorities for inter-regulatory co-ordination on overlapping concerns to promote regulatory harmony.
On enforcement, the CCI’s priority could be to prevent bigtech companies from gaining an unfair advantage by integrating their proprietary AI tools (e.g., Copilot and Gemini) into their existing products. Such bundling could reinforce bigtech’s dominance and create entry barriers for independent companies. The CCI should also keep an eye on any strategic acquisitions and partnerships in the AI space.
