This internship project aimed to improve epidemiological surveillance systems for Highly Pathogenic Avian Influenza (HPAI) through the analysis of data from online media sources. The main objective was to assess the performance of the PADI-Web tool in the early detection of HPAI outbreaks, by comparing its results to official notifications recorded in the WAHIS database. Using a multilingual corpus of media articles, events were identified, geolocated, and matched with official records based on temporal, spatial, and host-related criteria. The findings indicate that PADI-Web is capable of detecting certain outbreaks prior to their official reporting, although its coverage varies across different geographic and media contexts. This work highlights the strengths and limitations of automated event-based surveillance and suggests potential improvements to enhance its integration into official disease monitoring systems.