Desafíos actuales de la Inteligencia Artificial

58 Desafíos actuales de la Inteligencia Artificial 2.2.AI Regulation Under Competition and Intellectual Property Law Although the focus of this contribution is on exploring the interplay between the GDPR and the AI Act, it is also relevant to note that AI regulation is increasingly taking place in other fields, including competition and intellectual property law. The scope and objectives of all three areas differ substantially, though. The main objective of data protection law is to protect individuals’ privacy, and its scope is rather broad, as exemplified by the AI-related cases in France discussed in Section 4. Being technology-neutral and applying horizontally to both private and public entities, the GDPR applies to a wide range of practices involving AI systems. Its comprehensive rules and princi- ples leave sufficient room for interpretation, which is both its strength and weakness. Instead, the main objective of competition law in this context is to ensure that AI-inten- sive markets remain fair and contestable, guaranteeing a level-playing field for businesses, ultimately benefitting consumers. Here, trustbusters have to strike a difficult balance between innovation and competition on the merits. Traditional competition enforcement based on Articles 101 and 102 of the Treaty on the Functioning of the EU has led to important fines, also in AI-related markets. 13 Yet, traditional, ex post competition rules have been difficult to enforce in digital markets. Proceedings take years to result in sanctions and, in the meantime, harm to competitors and consumers is irreversible. In order to reverse this trend, the EU legislator adopted the DMA in 2022. Designed as an ex-ante regulation mainly concerned with ensuring competitive markets, the DMA complements traditional competition rules by applying more specific obligations in a number of defined digital markets controlled by gatekeepers designated as such by the Commission. While DMA enforcement is still in its infancy, some experts already argue that, con- trary to the AI Act, the DMA contains “the most far-reaching, most overlooked but poten- tially also most effective regulatory constraints for AI.” 14 What is sure is that the DMA will 13 For instance, the largest-ever competition fine of €4.34 billion was imposed by the European Commission on Google in 2018 (case AT.40099 – Google Android) – it was later reduced to €4.125 billion by the CJEU (judg- ment of 14 September 2022, Google Android, T-604/18, ECLI:EU:T:2022:541). The Commission had found that Google had abused its dominance in the market for mobile operating systems with Android to safeguard its dominance on the market for general search services, thus protecting its main source of revenue from its AI-driven search engine. 14 HACKER, Philipp; CORDES, Johann; ROCHON, Janina, “Regulating Gatekeeper Artificial Intelligence and Data: Transparency, Access and Fairness under the Digital Markets Act, the General Data Protection Regula- tion and Beyond”, European Journal of Risk Regulation (2023) 15(1), p. 51. At a recent workshop on generative AI, EU Commissioner for Competition Margrethe Vestager also confirmed that the DMA applied to generative AI features embedded in services covered by the regulation: “We continue to apply our trusty merger and anti- trust rules. Even though we sometimes have to remind AI market players that the competition rules also apply to them. And our Digital Markets Act applies too: the DMA can also regulate AI even though it is not listed as a core platform service itself. AI is covered where it is embedded in designated core platform services such as search engines, operating systems and social networking services. So we are applying our rulebook to concerns