AI and Antitrust: New Challenges for Competition Regulators

In recent years, artificial intelligence (AI) has transformed from a futuristic concept into a powerful tool embedded in the daily operations of businesses across industries. From dynamic pricing algorithms in e-commerce to predictive analytics in logistics, AI is reshaping how companies compete. But as AI becomes more sophisticated, it also raises complex questions for competition law. At the moment, regulators around the world are grappling with whether and how to adapt antitrust enforcement to this new technological landscape.

The Rise of Algorithmic Collusion

One of the most prominent concerns is the potential for algorithmic collusion – a scenario where AI systems, without explicit human coordination, learn to avoid price competition. Unlike traditional cartels, where companies agree behind closed doors to fix prices, algorithmic collusion can occur silently and autonomously. For example, pricing algorithms might observe competitors’ behavior and adjust prices in a way that stabilizes the market at supra-competitive levels.

This phenomenon, which is admittedly not that new (check out our blog-post from 2023), does not only challenge the traditional tools of antitrust enforcement but also raises questions as to the interpretation of the law. For instance, there may be no “smoking gun” emails or meetings to uncover. Instead, regulators must analyze (complex) algorithmic behavior and determine whether it can be categorized as purely legal observation of other companies or can somehow be seen as a meeting of “minds” and in turn as a violation of antitrust rules.

The Compliance Conundrum

For businesses, the intersection of AI and antitrust presents a compliance dilemma. On the one hand, AI can help firms optimize operations and respond more quickly to market changes. On the other hand, the same tools can inadvertently lead to anticompetitive outcomes. If, e.g., a large online retailer uses an AI-powered dynamic pricing algorithm to adjust prices based on inventory levels, customer demand, time of day and competitor prices (scraped from public websites), this allows the retailer to offer competitive prices, reduce overstock, and respond quickly to market changes – benefiting both the business and consumers. If in parallel, however, competitors in the same market use similar AI pricing tools from the same vendor or trained on similar datasets, the algorithms could converge on a pricing strategy that avoids undercutting which might be seen as a case of (tacit) collusion.

Companies must thus ensure that their AI systems are not only technically robust but also legally compliant. This requires a multidisciplinary approach. Legal teams need to work closely with data scientists and engineers to understand how algorithms function and how they might interact with competitors’ systems. Regular audits, algorithmic impact assessments, and clear documentation are becoming essential components of antitrust compliance programs.

AI as a Gatekeeper in Digital Markets

Another area of concern by regulators is how digital companies could use AI tools in their value chain. For example, some regulators seem to be concerned that large digital platforms might use AI to curate content, rank results, or recommend products and that this could have an impact on the competitive chances of other market participants.

The EU’s Digital Markets Act (DMA) does not cover AI/large language models as a distinct and regulated category of services, and there are already calls to broaden the DMA in that respect.

Regulatory Responses and Future Outlook

Regulators are not standing still. The attention paid to AI-driven market behavior, including potential collusion risks, shows the commitments of regulators to ensure effective competition in markets affected by AI and to monitor the competitive impact of AI.

An expert group meeting on AI and competition convened by the German Federal Cartel Office (FCO) this week shows this all too clearly (press release here). While the topics discussed do not come as a surprise (i.e., cloud services as an important part of the AI value chain; cooperations between AI and cloud service providers; development and deployment of foundation models; the importance of data as (scarce) input factor; use of AI applications in companies), it is questionable why noticeable that the expert group consisted exclusively of German stakeholders (Aleph Alpha, Axel Springer, Bosch, Brighter AI Technologies, Burda Media, GetYourGuide Deutschland, Innovate Europe Foundation, IONOS, the German AI Association, OpenGPT-X, Telekom Deutschland and Xayn (Noxtua)) and did not also include international players.

Some regulators are experimenting with algorithmic auditing – using AI to monitor AI. Others are calling for greater transparency and explainability in algorithmic decision-making. The debate continues over whether existing antitrust laws are sufficient or whether new rules are needed to address the unique challenges posed by AI.

Conclusion: A Delicate Balance

AI offers tremendous potential for innovation and efficiency, but it might also pose novel risks to fair competition. As we move deeper into the AI era, competition regulators try to strike a delicate balance: fostering technological progress while safeguarding market integrity. At the FCO expert group meeting, FCO President Mundt highlighted that “the Bundeskartellamt is […] keeping a close eye on the situation”. For businesses, the message is clear – AI is not a legal free zone. It must be developed and deployed with an awareness of antitrust implications.

The coming years will likely see a convergence of competition law, data governance, and AI ethics. Navigating this complex terrain will require legal expertise alongside technological literacy and a proactive compliance mindset.

Photo by Steve Johnson on Unsplash