Percepciones periodísticas de riesgos algorítmicos en comunicación electoral

This study examines how Colombian journalists with experience in electoral coverage perceive the risks and transformations introduced by artificial intelligence (AI) in political communication and democratic integrity. The aim is to analyze their assessments of automated content production, algorithmic influence on information visibility, and the institutional conditions that shape message circulation during elections in the Colombian electoral context. We used a qualitative–quantitative design, employing 20 semi-structured interviews processed with T-LAB. We applied cluster analysis and multiple correspondence analysis to identify semantic nuclei and interpretive axes. Findings show four main focal points: synthetic generation of disinformation, strategic use of algorithms to configure differentiated information exposure, reinforcement of polarization through affect-based automation dynamics, and digital security vulnerabilities linked to platform control. Two latent factors summarize these relationships: the transition from automated disinformation schemes to emerging frameworks of algorithmic oversight, and the shift from unprotected exposure environments toward increased institutional responsibility and operational transparency. While journalists develop systematic verification and lateral-monitoring strategies, these practices do not replace the obligation of State and corporate actors to regulate and audit algorithmic systems, ensure traceability of automated decisions, and uphold coherent regulatory mechanisms that protect democratic deliberation. This study underscores the need for strengthened institutional capacities and transparent governance to mitigate technological risks during electoral processes.