itinai content

←back to Blog

AI-Driven Antitrust and Competition Law: Algorithmic Collusion, Self-Learning Pricing Tools, and Legal Challenges in the US and EU

AI-Driven Antitrust and Competition Law: Algorithmic Collusion, Self-Learning Pricing Tools, and Legal Challenges in the US and EU

Understanding the Target Audience

The target audience for discussions surrounding AI-driven antitrust and competition law primarily includes legal professionals, business executives, policymakers, and academics. Their pain points revolve around navigating the complexities of new AI technologies in pricing strategies while adhering to existing antitrust regulations. They seek clarity on how AI impacts market dynamics and the legal frameworks governing competition. Moreover, they are interested in maintaining competitive advantage while ensuring compliance with evolving laws. Communication preferences lean towards detailed, data-supported analyses and updates on legal precedents, regulatory guidelines, and industry best practices.

AI in Market Economics and Pricing Algorithms

AI-driven pricing models, particularly those utilizing reinforcement learning (RL), can lead to outcomes resembling traditional collusion, fundamentally altering market dynamics. Unlike human-set strategies in oligopoly models, AI agents, such as Q-learning, autonomously learn pricing strategies from data, often resulting in supra-competitive pricing due to their ability to detect rivals’ actions and adjust in real-time. Such algorithms can mimic tacit collusion without direct coordination, often creating more stable, high-price outcomes than human actors could.

However, skepticism persists. In complex, noisy markets, economists argue that independent AI agents may struggle to form stable collusive strategies unless there is direct coordination, such as shared data. When AI-based coordination occurs via shared pricing data, it could violate antitrust laws. Algorithms often use large datasets to adjust pricing, and when non-public data is shared, it can subtly coordinate behavior.

Antitrust Law Perspectives

U.S. Law

Under the Sherman Act, price-fixing or conspiracies to restrain trade are prohibited. Courts require direct evidence of coordination, but using algorithms to coordinate pricing can still be seen as a violation if it results in cartel-like behavior.

EU Law

The EU’s competition law also prohibits anti-competitive agreements or practices under Articles 101 and 102 of the TFEU. If algorithms signal or align pricing systematically, it may be considered a concerted practice, akin to tacit collusion.

UK Law

Post-Brexit, the UK mirrors EU law and applies strict antitrust standards to algorithmic collusion. Algorithmic pricing without explicit coordination could still violate competition law.

Forms of Algorithmic Collusion

  • Explicit Cartels: Algorithms intentionally coordinate prices.
  • Tacit Learning Collusion: Independent AI agents autonomously settle on collusive pricing through self-learning.
  • Hub-and-Spoke Collusion: A third-party vendor’s software aggregates data from multiple firms to align pricing.
  • Algorithmic Signaling: Algorithms deduce rivals’ pricing from publicly available data and adjust accordingly.

Legal Frameworks

Predictable Agent Model

Firms are responsible for algorithmic behavior if they can predict and control pricing outcomes.

Digital Eye Model

If algorithms are highly autonomous and opaque, determining firm responsibility becomes more complex.

Graphical and Mathematical Models

Multi-agent reinforcement learning (MARL) underpins algorithmic collusion, where agents optimize long-term profits through repeated interactions.

Legal Challenges in Detecting and Prosecuting AI-Facilitated Collusion

Agreement and Intent

U.S. antitrust law under Section 1 requires proof of an intentional, concerted agreement. However, when AI agents independently learn from market conditions, no explicit agreement or human coordination may exist. In cases like Topkins, where direct communication occurred, collusion was clear.

Meeting of Minds for Non-humans

Traditional antitrust requires human agreement, but with AI, it’s unclear if an algorithm can “understand” collusion.

Mens Rea and Corporate Liability

While AI lacks criminal intent, liability can be ascribed to firms or human agents. Courts may treat AI behavior as the firm’s action, inferring liability if companies knew or should have known about their algorithm’s outcomes.

Evidence and Proof

Detecting algorithmic collusion is difficult due to the lack of traditional evidence like emails or meetings. Investigators might use circumstantial evidence to show intent.

Per Se vs Rule-of-Reason Analysis

There is ongoing debate about whether algorithmic pricing should be automatically deemed illegal. Courts in the U.S. and EU continue to wrestle with appropriate frameworks to assess competitive effects.

Regulatory Uncertainty and Enforcement Limits

Both U.S. and EU regulators face challenges in monitoring AI-driven markets. While studies on dynamic pricing and AI’s impact are ongoing, formal enforcement often begins only after substantial evidence emerges.

Enforcement and Legislative Responses to Algorithmic Collusion

Case Enforcement (U.S.)

  • Topkins (2015): The first criminal case against algorithmic price-fixing was recognized due to direct human coordination.
  • RealPage (2024): The DOJ filed against RealPage for enabling price-fixing in rental housing.
  • Duffy v. Yardi (2024): The use of RENTmaximizer was seen as potentially illegal price-fixing.

Regulatory Guidance and Private Enforcement (EU/UK)

  • EU: The European Commission has expressed concern over algorithmic collusion in its 2023 Horizontal Guidelines.
  • UK: The CMA has penalized Amazon resellers for illegal price coordination.

Legislative Efforts (U.S. and States)

  • PAC Act (2025): Would presume that exchanging sensitive information via pricing algorithms constitutes an agreement under the Sherman Act.
  • California Legislation (2025): California’s SB295 would criminalize certain uses of pricing algorithms.

Proposed Reforms (EU and Others)

  • EU AI Act: If passed, it would impose transparency and record-keeping requirements.
  • Global Coordination: The OECD advocates for international cooperation to address algorithmic coordination.

Proposed Reforms and Forward-Looking Frameworks for AI-Driven Collusion

  • Revisiting the Agreement Requirement
  • Algorithmic Transparency and Auditing
  • Enhanced Competition Compliance
  • Structural Remedies and Merger Review
  • Global Cooperation and Standards
  • Adaptive Enforcement Tools
  • Using Existing Tools

References

Calvano, E., Calzolari, G., Denicolò, V., & Pastorello, S. (2020). Artificial Intelligence, Algorithmic Pricing, and Collusion. American Economic Review, 110(10): 3267–3297.

Competition and Markets Authority (UK). Online sales of posters and frames (Case CE/98023). CMA Infringement Decision (August 2016).

European Commission. Guidelines on the application of Article 101 TFEU (2023), para. 379.

Giacalone, M. (2024). Algorithmic Collusion: Corporate Accountability and the Application of Art. 101 TFEU, European Papers: Insight 9(3), pp. 1048–1061.

OECD (2017). Algorithms and Collusion: Competition Policy in the Digital Age. OECD Publishing, Paris.

United States v. Topkins, No. 15-cr-00201 (N.D. Cal. Apr. 6, 2015).

United States v. RealPage, Inc., Case No. 1:24-cv-00710-WLO-JLW (M.D.N.C. 2024). DOJ Complaint (Aug. 23, 2024).

Duffy v. Yardi Systems, Inc., 64 F.4th 326 (9th Cir. 2023).

Competition Bureau Canada (2025). Algorithmic pricing and competition: Discussion paper (June 10, 2025).