Pricing Algorithms and Antitrust Law Concerns

Sidharth Chauhan, Harvard Law School

Synopsis

Anti-trust jurisprudence is based on a traditional understanding of contracts and arrangements. It depends on the evidentiary existence of arrangements or concerted practices of anti-competitive behavior. However, the increased usage of artificial intelligence (“AI”) for dynamic price-fixing of goods and products is challenging the contours of the anti-trust law regimes across the world. I am keen to explore how anti-trust laws will evolve in the future to deal with algorithmic price coordination amongst businesses without direct human intervention.

Background

Price-monitoring tools were initially developed to allow companies to track their competition and react immediately to price changes intended to ‘steal away’ their customers. These tools were expected to bring a higher degree of price transparency and intensify market competition. Businesses use these tools to efficiently and competitively offer their products to customers and make business decisions.[1] However, dynamic pricing tools may result in the facilitation of unregulated coordination among businesses, which may lead to cartel-like effects. The lack of human intervention makes AI-enabled algorithmic pricing a potential danger for the existing anti-trust regimes around the world.

Anti-trust concerns

‘Pricing Algorithms’ automatically sets prices with the objective of profit maximization. For dynamic pricing, such pricing algorithms typically have ready and spontaneous access to consumer data to track market conditions on a real-time basis. They enable businesses to react immediately and implement continuous price changes.[2]

The availability of real-time consumer data give these algorithms the extraordinary ability to perceive, develop and act on this information for adjusting prices for adapting to the changing market conditions. Since human intervention need not be involved in an AI environment, any unregulated machine-led pricing coordination adds an additional level of vulnerability to a competitive market.[3]

It is possible that algorithmic pricing could create a market setting in which competitors can engage in a kind of automated price fixing without the risk of being held liable under competition law.[4] No agreement or communication is necessary between competitors for their algorithms to engage in interdependent pricing. This would lead to anti-competitive effects without human collusion or intent.

EU legislative framework

It is interesting to explore whether the current competition law enforcement tools are sufficient to combat AI-enabled pricing concerns.

Articles 101 and 102 of the Treaty on the Functioning of the European Union (“TFEU”), which prohibit anti-competitive price-fixing within the EU, do not specifically recognize automated price-fixing through the use of algorithms. In the case of Bayer AG, the Court of First Instance of the European Communities has stated that Article 101 of the TFEU requires “the direct or indirect finding of the existence of the subjective element that characterises the very concept of an agreement” as well as “a concurrence of wills between economic operators on the implementation of a policy, the pursuit of an objective, or the adoption of a given line of conduct on the market, irrespective of the manner in which the parties’ intention to behave on the market in accordance with the terms of that agreement is expressed.[5] If price-fixing through AI is the result of a conscious human decision, such practice should fall within the scope of Article 101 of the TFEU. However, the present EU anti-trust jurisprudence does not cover simple AI-driven pricing decisions where there is no agreement or collusion between parties (explicit or implicit). In fact, some commentators are of the view that AI could be used to facilitate all types of transactions that are otherwise prohibited under Articles 101 and 102 of the TFEU.[6]

Conclusion

Anti-trust authorities need to be given the power to analyze the nature of algorithms, their effects on markets, and the extent of human involvement while ascertaining the impact of AI in anti-competitive behavior. Since profitability is a factor taken into consideration by AI to determine spontaneous pricing, it is likely that the AI algorithms being utilized by competing businesses will settle on a ‘pricing comfort zone’ resulting in a ‘mathematical collusion’ between profitability and pricing at macro levels.

Until now, actions by the European Commission and the Federal Trade Commission have been in cases where human collusion was involved, albeit through the use of AI tools. For instance, in the Topkins[7] case, competing online sellers agreed not to undercut each other’s prices and implemented this agreement through ‘automated repricing software’ that automatically set the costs of the products to stay in line with other online sellers. Similarly, fines of €111 million were imposed upon on Asus, Denon & Marantz, Philips and Pioneer for fixing the minimum resale prices of their online retailers. The four manufacturers impeded online retailers from offering their products for prices that were lower than those recommended by the manufacturers. Price monitoring tools were used to orchestrate resale price for such extended periods and across numerous territories.[8]

Such non-human collusion may have serious economic and anti-competitive effects on the society. The present anti-trust regime needs to recognize these challenges and identify sufficient checks and balances to ensure that machine-driven dynamic pricing models do not lead to an anti-competitive environment.


The author is an alumnus of Government Law College, Mumbai and is currently a working professional dealing with private equity, mergers and acquisitions. He will be joining Harvard Law School this fall. His LinkedIn profile can be accessed here. 

Views are personal.


Notes

[1]     Note. From “Robots in law: How artificial intelligence is transforming legal services” J. Goodman, 2016. 

[2]     See OECD, (2017) Algorithms, and collusion: Competition policy in the digital age, page 9.

[3]     Note. From “The MIT encyclopedia of the cognitive sciences “ R. Wilson, & F. Keil, 1999, page 31.

[4]     Note. From A. Ezrachi and M. Stucke, “Virtual competition, The promise and perils of the algorithm-driven economy,” Harvard University Press, November 2016, chapters 7 and 8.

[5]     See (2000) Bayer AG v Commission of the European Communities, T-41/96, § 69.

[6]     Note. From A. Krausová “EU competition law and artificial intelligence: reflections on antitrust and consumer protection issues” 2019. https://tlq.ilaw.cas.cz/index.php/tlq/article/download/322/321.

[7]     See. (2015) United States of America v. Topkins, No. 15-00201 WHO (N.D. Cal.).

[8]     Note. From Reuters “EU Antitrust E-commerce Fine”; available at https://in.reuters.com/article/eu-antitrust-ecommerce-fine/eu-fines-philips-asus-pioneer-denon-marantz-total-111-million-euros-idINKBN1KE1GH; accessed on 13 June 2020.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s