When Bots Set Prices: CMA Highlights Real-World Risks of Algorithmic Pricing
Background
In a recent blog post, the UK Competition and Markets Authority (CMA) highlighted the increased antitrust compliance risk attached to the use of AI-driven algorithmic pricing tools. These tools have become increasingly powerful and widespread, delivering significant operational efficiencies and commercial benefits. However, they may also heighten the risk of collusion between competitors, prompting the CMA to focus on potential breaches of UK competition law.
Similarly, the European Commission has identified algorithmic pricing as an enforcement priority, emphasizing the competition law risks in its Horizontal Cooperation Guidelines published in 2023. This scrutiny is part of a broader regulatory landscape aimed at addressing unfair technology-driven pricing practices, such as ‘drip’ and ‘dynamic’ pricing.
Rise of Algorithmic Pricing: Why is the CMA Paying (More) Attention?
Algorithmic pricing is not a new phenomenon; it has been utilized across various sectors, including air travel, hospitality, and retail, for decades. However, the sophistication and ubiquity of these pricing algorithms have markedly increased. Modern algorithms process large-scale datasets in real-time and are increasingly powered by cutting-edge large language models (LLMs), providing businesses with unprecedented access to powerful predictive technologies.
The CMA has also published a paper on agentic AI, which identifies the competition risks associated with autonomous agents used to optimize pricing strategies. Interactions between agents from competing businesses may reduce competitive pressure, leading to agentic collusion.
Algorithmic Collusion: Heightened Compliance Risks for Businesses
The CMA outlines several ways in which AI incorporation into pricing algorithms can lead to coordinated anti-competitive outcomes:
- Implementation of ‘classic’ collusion: Competitors may agree to coordinate their commercial conduct and use algorithms to enforce their agreement.
- Hub-and-spoke collusion: Competitors may share the same algorithm or data hub to exchange sensitive information indirectly.
- ‘Predictable agent’ behaviour: Algorithms that react predictably to market events can lead to tacit coordination, softening competition.
- Autonomous AI coordination: Advanced AI may autonomously learn to reach coordinated outcomes aimed at maximizing profits.
The CMA’s concerns are substantiated by a growing body of academic research and empirical studies indicating an increased risk of algorithmic collusion. Recent enforcement actions by the CMA and regulators in Europe and the U.S. further underscore these compliance risks.
Recent Enforcement Action
The deployment of algorithms has attracted scrutiny across various sectors, reaffirming that the use of algorithms does not exempt businesses from antitrust violations. Notable cases include:
- UK Online Retail: The CMA found that two online sellers of posters engaged in anti-competitive practices using automatic repricing software.
- Hotels: An investigation was launched against leading hotel chains suspected of sharing sensitive information via a hotel data services provider.
- Italy Air Travel: The Italian competition authority explored the impact of pricing algorithms, indicating a need for improved price transparency in airline ticketing.
- EU Online Travel: The European Court of Justice ruled that travel agents could be liable for anti-competitive collusion facilitated by shared algorithms.
- US Meat Processing: The DOJ filed a lawsuit against Agri Stats for organizing the exchange of sensitive information between poultry processors.
Practical Takeaways for Businesses: How to Mitigate Compliance Risks
While the use of algorithms or AI is not inherently problematic, businesses must be aware of the potential compliance risks involved. To mitigate these risks, companies should take several proactive steps:
- Due diligence: Vet new algorithmic pricing tools during procurement to understand their operation and underlying data sources.
- Governance and oversight: Establish clear policies governing the use of these tools, including record-keeping and manual overrides.
- Audits: Regularly audit AI tools, including data inputs, use cases, and employee access.
- Compliance training: Educate employees about the legal risks associated with pricing algorithms and the exchange of sensitive information.
In conclusion, as algorithmic pricing tools continue to evolve and proliferate, businesses must navigate a complex landscape of regulatory scrutiny and compliance risks to avoid potential pitfalls.
