Introduction
The recent advent of Artificial Intelligence (AI) into the mainstream of every field– including usage by the government in aid of administration, means that many AI-developing contenders have shown up on the horizon. With the AI market developing rapidly and branching off into sub-categories just as swiftly automated artificial intelligence mechanisms now cater to almost every imaginable need, not only recreational and domestic, but also institutional, industrial and enterprise. As the AI industry comes close to growing into a formalized industry, it becomes paramount to understand the effect it will have on competition.
To understand the competitive potential of Artificial Intelligence, as well as any detrimental effects to competition and finally, its transformative capabilities, the Competition Commission of India (CCI) will launch a “Market Study on Artificial Intelligence (AI) and Competition” in 2024. It opened invitation for proposals on April 22, 2024, and accepted the last batch on June 03, 2024.
The objectives of the Market Study are to:
– Understand key AI systems and markets – stakeholders, essential inputs, value chains, market structures and parameters of competition;
– Examine potential competition issues, if any;
– Study the scope and nature of uses and application of AI, and use the information to assess risks, opportunities and implications from the perspective of competition;
– Understand the regulatory framework – both existing and evolving – governing AI in India and other major jurisdictions;
– Reach out to all relevant stakeholders and discuss the intersectionality between AI and competition; and
– Understand AI trends and prioritize the relevant enforcement and advocacy steps to be taken by the CCI with regard to AI and the markets
The pro-competition application of AI in user industries includes the usage of AI for purposes such as monitoring, recommending, and predicting trends in various arenas such as retail, e-commerce, marketing, etc. Conversely, the concerns that arise include the new forms of collusion that may arise due to the use of self-learning AI mechanisms, as well as abuse of dominance. Apart from that, the influence of AI over entry and exit barriers and market concentration of firms is also uncharted territory. The lack of robust legal/regulatory framework in India also leaves a gap behind.
Areas of Intersection
The relation between Artificial Intelligence and competition law is one with a synergistic character. There is a confluence between the pro-competitive market effects of AI implementation and the anti-competitive market effects. The anti-trust authorities may themselves equip AI mechanisms and make use of them for the detection of violations and contraventions of competition law. These benefits and threats to competition posed by AI presents opportunities and challenges for authorities to consider and safeguard competition.
Pro-Competitive Implications
Employing the use of AI, directly or indirectly, has the power to lead to pro-competitive benefits on both sides of the market – the supply side as well as the demand side. AI provides a variety of goods through its vast database, tailors them to consumer-needs, boosts supplier-reach, and increases attractiveness of the product on the digital shelf (in terms of its availability, prices, quality, and innovation). In more ways than one, it can be said that AI contributes to consumer welfare and fulfils one of the most paramount objectives of competition law, if not its ultimate goal.
Innovation and Product Development
Due to the wide datasets available to AI, as well as its predictive analysis capabilities, coupled with its advanced algorithmic functions, AI is an R&D asset. This is to say that it makes knowledge more available, facilitates market trend predictions and aids in risk-estimation of R&D projects, thus, inspiring the innovation of new products.
It accelerates the research and development process by providing synthesized data, both on the market and the consumer, as well as on the product and its demand. This also enables businesses to provide personalized products and meet the consumer’s diverse and specific requirements via personalized search results. This may also conversely result in an anti-competitive strategy where the consumer gets unknowingly nudged into preferring one product over the other due to the AI-engineered “personalized” feed. For example- a website may use eye-catching display features, such as huge fonts, to encourage specific search results over others, or they may alter the order in which the results appear on screen to push a few selected ones on top, psychologically nudging the consumer to make choices out of a pool that may not be as unbiased as assumed by him.
Lowering Barriers to Entry
AI lowers barriers to markets by making them more transparent. The easy availability and accessibility of AI tools has opened the arena for small firms and startups, thereby lowering market barriers. Another one of the reasons for such lowering is the reduction of cost of production and operational costs, which increases the number of active participants in the market, and boosts competition.
AI platforms provide consumers with optimized decision-taking tools with access to commercial information and protection from manipulative market techniques, including price discrimination, also increase the transparency of the market environment. This improves the flow of commercial information which may encourage people to choose different supplies, even small firms or startups, thus lowering entry barriers, and encouraging more participants into the competitive markets.
Consumer Benefits
AI has given new tools to consumers to take specific decisions, alterable to their specific needs and requirements. It uses the vast consumer and product dataset available, and upon input commands by the user recommends personalized search results and transactions that cater to every user’s individual and business needs.
AI tools can also be used for the purpose of guiding consumers on market prices and conditions. It also allows for a significant reduction of costs to be born, both by the consumer and the supplier.
It may also lead to a new set of consumers – the “algorithmic consumers” whose regular decision-making is dependent upon the application of algorithms, wholly or partially to say the least.
Dynamic pricing
Price algorithms used by AI respond to changes in the demand and supply conditions, and instantaneously recommend fluctuating prices to meet the conditions. This leads to optimization of prices, or dynamic pricing. AI enables these dynamic pricing models that also lead to the optimization of resource allocation, reduction of waste, and even promotion of sustainable business practices, thus, enhancing market efficiency.
Anti-Competitive Ramifications
Predictability, transparency and frequent engagements among the algorithms of market competitors are some of the challenges faced by competition due to the boom of the AI industry. For example- price transparency on one hand aids the consumer to make an informed decision, on the other hand it may lead to price co-ordination among firms which adversely affects competition. The predictability of the common algorithms used by competitors and their frequent interactions undermine the unique-character of each firm which paves the way for price-collusion. Further, AI also has the potential to give rise to new competitive concerns, namely, those specific to digitalized markets.
Anti-Competitive Strategies
AI might opt for anti-competitive strategies when it is made the driver of competitive decisions about, say, prices or quality in a dominant firm. For example- machine-learning (ML) algorithms with long-term sight might opt for predatory-pricing even in absence of instructions to do so. Predatory pricing refers to a strategy where a firm cuts prices to push all competitors out of the market, and then subsequently increases the prices upon having obtained the desired market power.
The use of AI in online commerce platforms that connect the firm to the consumer may also lead to anti-competitive practices such as “self-preferencing”, i.e. where a firm pushes its own products in the search results, or “nudging” where the AI shows prominent display features in an attempt to encourage a specific choice while the consumer believes the choices presented to him via the search results to be neutral or unbiased.
AI might also lend a hand in encouraging the practice of “personalized pricing” wherein price discrimination occurs by charging different consumers with different prices based on their consumption patterns. AI pushes it in the way that firms can process wide datasets on consumer characteristics which allows them to set individually-tailored prices.
Machine-learning AI in particular is prone to giving unexpected or incorrect results due to a variety of limitations such as algorithmic bias or bias in the training data which is inevitable due to the human-aspect involved which introduces their personal qualifications, inherent biases, and their considerations of compromises and trade-offs. Another limitation is the mutual exclusivity of its accountability and transparency – there is generally a trade-off between the accuracy and transparency of an AI model due to its “black box” nature wherein it cannot always explain the logic behind its conclusion.
Monopolistic concerns
AI is a powerful tool that may bring about monopolistic concerns wherein large firms might try to create a monopoly and abuse their dominance. It could become a distinctive differentiator among firms in terms of who has access to the latest technologies, with the more accurate algorithmic mechanism. Firms will also be differentiated in terms of their abilities to quickly respond to market changes, to accurately predict and interpret data, and to harness AI as a tool. On the other hand, all these differences may even lead to the contrasting conclusion of increased competition, depending on the nature of the market.
Price fixing and Collusion
While price transparency brought about by AI aids consumers in their decision-making processes, it may also dampen competition by incentivizing firms to coordinate and collude, leading to price-fixing.
Collusion refers to a form of coordination or agreement among competing firms to raise profits higher than those obtained at a non-cooperative equilibrium. As there is a high degree of market transparency owing to the wide availability of data, it is easier for firms to collude and then communicate through price-signals to keep the collusion intact. Collusion includes agreements towards price fixing, production quotas, market-sharing, and bid-rigging etc. that are anti-competitive. Their purpose is to control the market and eliminate competition by collectively raising prices, restricting outputs, manufacturing scarcity and creating a joint-monopolistic market where all colluding members profit, and the consumer suffers. This is detrimental to the competitive spirit.
Besides, larger AI firms are likely to compete in multiple markets by being utilized by conglomerate business models. Research compiled by the OECD suggests that in such a situation where competitors may be in contact with each other across multiple markets, collusion is more likely.
Breach of Consumer Data Privacy
On one hand, common availability of data hinders competition through collusion, on the other hand, disproportionate access to or control over consumer data can adversely impact competition by creating a monopolistic market. AI may be used to track user search patterns, application history, and individual consumption trends via surveys, advertisements, tracing clicks and buys on e-commerce websites, and the time spent by the user on each website. All this data can be easily tailored to uncover hidden patterns and used to generate personalized products or services and target consumers based on these specifications.
This is possible because AI models in addition to producing descriptive output, can also predict future consumption behavior by its predictive analysis usage. Thus, AI in this way also redefines the assumption that the average consumer has the ability to make autonomous decisions. Bigger firms may have access to better technologies which allows them to breach consumer privacy in a way that goes unnoticed by said consumers, thus, creating an imbalanced advantage in their favor, which in turn is unfavorable to competition.
Conclusion
The understanding of AI’s implications on competitive dynamics is still in its infantile preliminary stages. It is yet more hypothetical than factual. AI has some pro-competitive character to itself wherein it leads to consumer benefits, enhanced compliance, dynamic pricing, lowering of entry barriers and accelerated innovation. On the other hand, AI also holds the potential power to dampen competition by making markets predictable, static, monopolistic and transparent.
In order to improve the community’s understanding of these ramifications, and subsequently lay down the foundation for legal/regulatory framework for the same, some tools may be deployed by competition authorities. These tools include market studies, merger control, co-operation among regulating authorities, and competition advocacy.
Although regulation is the instinctual reaction to every new technological evolution, the uncertainty and skepticism bound to Artificial Intelligence calls for caution against following the siren’s call that asks for more and more regulations with each innovation. Rushing into a regulatory framework for Artificial Intelligence is more likely than not to be premature and counterproductive at a stage where the multifarious facets of AI are yet unexplored. Therefore, the CCI’s decision to hold a market study is a prudential plan at present.
Author(s)
Kanika Sharma
Student at AIL, Mohali