User Autonomy and Algorithmic Recommendations under the Digital Services Act, the Digital Markets Act, and the Political Advertising Regulation

The paper analyses the advancements and limitations of certain recent regulations on digital services in protecting the autonomy of citizens online, particularly concerning content transmission and reception. Recommender systems and algorithms play a central role in determining the content users encounter online, including on social media platforms, search engines, and generative AI systems. However, algorithmic content transmission presents significant challenges for values, due to its opaque functioning and its reliance on a vast array of personal data, increasing the risk of manipulation. Recent regulations, broadly designed to place citizens at the centre of the digital environment, seek to address these concerns through three key mechanisms (a) specific transparency obligations, (b) reinforced freedom of consent, and (c) prohibition of most high-risk practices. While these new rules represent an important step toward strengthening user autonomy in relation to content recommendations, their effective implementation remains questionable, and they remain limited in providing users with mechanisms to actively shape recommendation parameters.