Politics or People? An AI-Assisted Media Framing Analysis of the 2025 Pakistan Floods Across Local and International Broadcast News Media

Politics or People? An AI-Assisted Media Framing Analysis of the 2025 Pakistan Floods Across Local and International Broadcast News Media

Authors

  • Dr. Muhammad Aftab Madni Assistant Professor, Department of Media and Communication Studies, Shaheed Benazir Bhutto University (SBBU)

DOI:

https://doi.org/10.59075/jssa.v4i1.611

Keywords:

Pakistan floods; media framing; AI-assisted content analysis; disaster communication; broadcast journalism; climate change; agenda-setting; NLP; political framing; humanitarian reporting

Abstract

Despite the scale and recurrence of such disasters, empirical research systematically comparing domestic and international broadcast framing of Pakistan's floods remains sparse. This study addresses that gap through a comparative, AI-assisted content analysis of 400 flood-related television news stories sampled from four major Pakistani broadcasters (Geo News, ARY News, Samaa TV, and Dunya News) and four international outlets (BBC World News, Al Jazeera English, CNN International, and Deutsche Welle), covering the period July 15 to September 15, 2025. Drawing on framing theory (Entman, 1993) and agenda-setting theory (McCombs & Shaw, 1972) as its conceptual anchors, the study systematically codes each story for dominant frame (political/governance, human impact, relief/response, economic/infrastructure, climate/environmental), tone (empathetic, neutral, blame-oriented), and actor visibility (politicians/elites, citizens/victims, NGOs/INGOs, experts/scientists, military/rescue officials, international donors). Natural language processing (NLP) tools—including OpenAI Whisper for speech-to-text transcription of Urdu-language bulletins, neural machine translation, named entity recognition (spaCy), transformer-based sentiment analysis, and BERTopic for topic modelling—are integrated with manual coding to ensure scalability, consistency, and bilingual coverage. Findings reveal that Pakistani broadcasters relied predominantly on political and governance frames (60%), foregrounded politicians as primary actors (50%), and adopted blame-oriented tones (40%), while international outlets prioritised human impact (40%) and relief frames (20%), employed empathetic storytelling (45%), and gave greater visibility to citizens, NGOs, and humanitarian experts. Topic modelling and sentiment analysis corroborated these patterns, confirming an elite-centric news culture in domestic coverage that may privilege political spectacle over humanitarian urgency. These findings carry significant implications for journalists, policymakers, and humanitarian actors seeking more equitable, people-centred disaster communication in Pakistan and across the Global South.

Downloads

Published

2026-03-17

How to Cite

Dr. Muhammad Aftab Madni. (2026). Politics or People? An AI-Assisted Media Framing Analysis of the 2025 Pakistan Floods Across Local and International Broadcast News Media. Journal for Social Science Archives, 4(1), 1763–1777. https://doi.org/10.59075/jssa.v4i1.611
Loading...