Safeguarding media freedom from infrastructural reliance on AI companies: The role of EU law

An emerging body of journalism studies research has shown how media organizations are growing dependent on external companies to provide AI tools used to inform the public, and the infrastructure needed to develop and deploy these tools. Concurrently, EU lawmakers and legal scholars have developed new regulatory and normative frameworks to safeguard media freedom from large technology companies. However, this work focuses on platforms' control over access to large audiences; it remains unclear how AI companies' power over infrastructure inside newsrooms challenges media freedom. This paper therefore explores how European law should address the challenges to media freedom posed by the media's dependence on the infrastructure controlled by AI companies. It does so in two steps. First, it evaluates why the media's dependence on AI companies poses a challenge to the fundamental right to media freedom. It finds that media organizations' loss of control over the values embedded in the AI tools they use to inform the public poses the most pressing challenge. Second, it explores the suitability of existing EU law to address three conditions (algorithmic opacity, lock-in effects, and resource disparities) for the media's infrastructural reliance on AI companies. It finds that existing EU law does not adequately address these conditions. However, especially horizontal regulation targeting AI tools and the underlying cloud infrastructure do offer regulatory tools that can be applied or adapted to safeguard media freedom from infrastructural reliance on AI companies.