Recommendation Algorithms: how AI shapes our media preferences

Abstract

Problem Statement. The rapid implementation of recommendation algorithms in digital media is radically transforming the structure of information consumption, shaping individualized media preferences of users. Personalization enhances convenience and audience engagement, but simultaneously intensifies the risks of information bubbles, manipulation, asymmetric data access, and reduced transparency in editorial policy.

Purpose/Objectives. The purpose of this article is to comprehensively analyze how AI-based recommendation algorithms shape users’ media preferences, to identify their impact on editorial strategies and the information landscape, and to outline key ethical, social, and regulatory challenges. The objectives are: to describe the main types of algorithms and data sources, to examine their influence on audience behavior and media business models, and to analyze risks to pluralism and trust.

Methods. The research is based on a combination of content analysis of scientific and industry publications, case analysis of global and Ukrainian online publications, as well as elements of comparative analysis of regulatory approaches to the use of AI in media. A structural-functional approach was applied to study the role of algorithms in the media system, along with critical discourse analysis regarding the ethical dimensions of personalization.

Results/Conclusions. It is demonstrated that recommendation algorithms based on collaborative and content-based filtering, as well as hybrid machine learning models, are becoming a central element of interaction between media and audiences, determining content visibility and news consumption trajectories. They contribute to increased engagement and monetization, but simultaneously narrow users’ information horizons, intensify polarization, and complicate the separation of editorial decisions from automated ones. The conclusion emphasizes the necessity for transparency in algorithmic principles, labeling of personalized content, implementation of ethical standards in media organizations, and strengthening audience media literacy.

Novelty. The scientific novelty lies in examining recommendation algorithms in media not merely as a technical tool for optimizing user experience, but as a complex sociotechnical mechanism that reshapes the structure of the public sphere. The article combines analysis of the technological principles of AI operation with an assessment of their impact on editorial autonomy, media pluralism, and democratic processes, enabling a holistic approach to developing ethical and regulatory frameworks for the use of recommendation systems in journalism.