What compliance steps do businesses need to consider as regulatory scrutiny rises?

Antitrust authorities are increasingly  alert to the risk of algorithmic pricing generating anti-competitive conduct. The recent OECD report on algorithmic pricing and competition providing an overview of the work of the G7 antitrust authorities in this field is a further sign of this trend. So far, actual enforcement remains limited, except in the US. Relying on recent public statements by European regulators, this may be about to change.

While the G7 regulators have identified convergent concerns, the report notes that the extent and means of enforcement remain globally varied. We review the areas of concern, before outlining steps that businesses deploying algorithmic pricing systems may consider to minimise antitrust risk.

What is the Concern?

The authorities have examined the potential anticompetitive risks that pricing algorithms pose for both coordinated behaviour (i.e., cartels) and unilateral conduct (i.e., abuse of dominance). They focus on specific theories of harm, namely: 

  • Collusion. Algorithms may (i) facilitate “traditional” anticompetitive pricing agreements; (ii) enable hub‑and‑spoke collusion in which a common algorithm provider serves as the “hub” connecting competing “spokes”, leading to alignment of pricing and/or information flows; and (iii) foster tacit coordination through autonomous learning, where separate algorithms deployed by competitors begin to align on pricing.
  • Unilateral Conduct. Pricing algorithms, particularly when seeded with extensive data, can enable sophisticated pricing strategies. While algorithmic price discrimination and personalisation may deliver efficiencies and tailored offers, antitrust agencies warn that they also heighten the risk of exploitative outcomes and conduct with exclusionary effects.
  • Consumer Concerns. Algorithmic techniques such as dynamic pricing also attract concerns regarding transparency, consumer trust and discriminatory pricing. These consumer law-tinged concerns often colour the framing of traditional antitrust cases, as well as giving rise to consumer law cases in their own right (see the Ticketmaster case in the UK).

Increasing Scrutiny

In recent months, competition authorities have launched a range of initiatives to deepen their understanding of pricing algorithms, including pursuing enforcement actions, conducting surveys, and undertaking market studies.

  • US Developments. The US DOJ’s lawsuit against RealPage (a revenue management software provider, acting as the “hub”) and a number of property management companies (the “spokes”) alleges the defendants entered into unlawful agreements to exchange competitively sensitive information and to achieve pricing alignment, keeping rental prices high. Private plaintiffs have brought parallel class action lawsuits against RealPage and the property management companies, as well as challenges of pricing software used by hotel operators. Algorithmic pricing also received attention in the reported use by US airlines of AI-based software provided by Fetcherr to set fares. California has amended its antitrust law (as our client memo explains) to clarify that price-fixing coordination is unlawful whether orchestrated through traditional means or implemented via algorithms.
  • Enforcement Priorities Across Jurisdictions.
    • Regulators in Europe are also focusing on algorithmic pricing. Market studies and public consultations make this a clear priority, and European Commission officials have confirmed that it is investigating (still confidential) algorithmic pricing matters. The Commission is planning a study on “modern pricing and the risks of collusion”, which may well discuss algorithmic pricing and revenue management systems. At the national level, the Dutch and Italian authorities have opened investigations into airlines, which have long used algorithmic techniques in pricing. The UK has also examined dynamic pricing driven by algorithms in a study covering several industries, finding both benefits and drawbacks.
    • The Japanese FTC is monitoring the use of algorithms to set low prices only to competitor’s customers to hinder competition, possibly triggering a unilateral violation of competition law.
    • The Canadian Competition Bureau opened an investigation into Kalibrate, a pricing and analytics software provider, to determine whether its pricing services are lessening competition and enabling collusion among gas stations in Canada.
  • Evidentiary Challenges.
    • Authorities worldwide acknowledge evidentiary hurdles, particularly for tacit algorithmic coordination. In the US, an agreement has to be proved. In the EU, a concerted practice suffices. Antitrust rules on anti-competitive agreements do not require a formal agreement but require a “meeting of minds” between parties. Unlike traditional cartel conduct that may generate meeting minutes, messages or other documentary traces, algorithmic interactions often leave limited or no record. In fact, if the collusion has been instigated by interactions between self-learning algorithms, the parties may not even be aware that it is happening.
    • In the US, DOJ and private plaintiffs have argued that the agreement element is met when a provider of a pricing algorithm offers a common algorithm to competitors who each accept the offer (i.e., sign up for the service) knowing that their competitors are doing the same.
    • It can be difficult for authorities to find evidence to demonstrate communications or any manifestation of mutual commitment, if such commitment exists at all. Therefore, authorities rely on market studies and monitoring initiatives to gather data and identify patterns consistent with coordinated outcomes.

Practical Takeaways for Companies – Risk Mitigation

To mitigate the likelihood of regulatory scrutiny, companies using algorithmic pricing systems should consider the following steps to mitigate compliance risks:

  • Clean data practices. Exercise care when seeding algorithms with market sensitive data like pricing, strategy, capacity, stock levels etc. If the algorithm is used by competitors, then only publicly available data should be used.
  • Consider risk from pricing management software vendors. Audit the risk associated with third-party vendors of applications that support pricing and revenue management decisions. Companies should ensure that third-party vendors keep their data separate from the data of other companies. But companies should also consider the broader antitrust risk associated with such systems, particularly where pricing authority is delegated to and/or confidential firm information is being shared with, or received from, the vendors of the system.
  • Provide training. Key stakeholders from any team that is involved in deploying algorithm-based pricing or revenue management applications should be included in routine antitrust training so that they can spot issues and escalate them appropriately.
  • Obtain advice. As agencies’ data units mature, companies should expect more frequent requests for information, deeper engagement with expert evidence and agencies’ scrutiny. Legal teams should be informed of major pricing product tools or developments and given an opportunity to comment on risks before they are adopted. Because agencies currently provide little clear guidance and the enforcement landscape is evolving rapidly, seeking early legal advice is essential to prepare in advance pro-competitive arguments and anticipate scrutiny by agencies.

 

* * *