用AI分析2025年以伊冲突视频评论，发现数字舆论偏袒巴勒斯坦

Purpose. This study analyzes the digital representation of the Iran-Israel conflict that occurred in June 2025, based on 120,000 comments posted on YouTube. It sought to identify discursive positions regarding the actors involved and to examine how media and algorithmic biases shape digital conversations. Methodology. A mixed-methods design with triangulation was adopted. In the quantitative phase, natural language processing techniques and machine learning models (BERT and XLM-RoBERTa) were used to classify comments into ten categories. In the qualitative phase, a critical analysis of media context and ideological narratives was conducted, complemented by manual annotation and supervised training. This strategy enabled the integration of statistical robustness with contextual understanding. Results and conclusions. The findings reveal a clear overrepresentation of pro-Palestinian and anti-United States/Israel discourses, while pro-United States and anti-Palestinian positions were marginal. Iran, usually rendered invisible in global media, emerged as a central actor in the digital conversation during the conflict, suggesting a narrative shift away from previous hegemonic frameworks. Likewise, the results confirm the influence of algorithmic biases in amplifying certain discourses while limiting others. Original contributions. This work combines computational analysis and philosophical critique for the study of digital controversies, providing a methodological framework replicable in geopolitical contexts. It is one of the first Spanish-language studies to map, through artificial intelligence and critical analysis, discourses on an international conflict on YouTube, highlighting asymmetries and narrative disputes that are often overlooked.