EU Trends and Developments Contributed by: Andrzej Kmiecik, Richard Burton and Catherine Gordley, Van Bael & Bellis
Artificial Intelligence: Policing Uncertain Boundaries Promoting technological innovation is regarded as one of the key paths for the EU to boost its competitiveness while meeting its sustainabil - ity goals. A particular area where technological innovation may collide with the cartel prohibition is the increased use by companies of artificial intelligence (AI) and algorithms in plotting their commercial strategies. While the rapid evolution of AI has been recognised as one of the most important developments of recent years and as a potential catalyst to boost innovation and economic growth, antitrust enforcers in Europe and elsewhere in the world have also expressed concerns about the possibility of the increasing use of AI tools (such as pricing algorithms) lead - ing to anti-competitive outcomes. While so far there has been next to no precedent in this area, in its 2023 Horizontal Guidelines, the Commission gave guidance on how the antitrust rules can be applied to algorithmic pricing tools, identifying two scenarios where they are likely to constitute cartel-like behaviour: • so-called “collusion by code” , where compet - itors agree on the use of a common pricing algorithm; and • hub-and-spoke situations, where competitors knowingly use the same third-party pricing tool, and this leads to unlawful exchanges of competitively sensitive information between them. By contrast, the Horizontal Guidelines suggest that independent use of an algorithmic tool that uses publicly available data would in principle be lawful. In practice, while detecting “collusion by code” would appear relatively straightforward, draw -
ing the boundaries of unlawful conduct where competitors use the same pricing third-party tools may be more challenging, and may involve an assessment of evidence of awareness and foresight among the companies concerned that leaves considerable uncertainty. It seems likely that the Commission will seek to step up its car - tel enforcement efforts here, which would also set a welcome precedent clarifying the applica - tion of the cartel prohibition in this complex and fast-moving area. However, whatever further clarity emerges from future precedent, it is likely that the use of AI tools in pricing and commer - cial decision-making will require companies to always carefully assess possible cartel compli - ance risks. Aside from this, the broader question is wheth - er, given the speed at which AI is developing, enforcement policy will be able to adequately keep pace. Apparently aware of these challeng - es, in a July 2024 joint statement on competi - tion in generative AI foundation models and AI products, the Commission, along with the US Department of Justice, the US Federal Trade Commission and the UK Competition and Mar - kets Authority, emphasised its commitment to monitoring and addressing specific risks that may arise as part of their broader efforts to ensure that the AI ecosystem remains competi - tive. It remains to be seen how successful the Commission and other antitrust enforcers will be in rising to these challenges. Price Signalling: Cartels in Plain Sight While the growth of AI presents challenges to cartel enforcement through the possibility of collusive outcomes being generated in IT server rooms without human intervention, at the other end of the spectrum, the Commission has also recently spelled out its intention to crack down on what might be considered cartels operating
91
CHAMBERS.COM
Powered by FlippingBook