Media Coverage of War Victims: Journalistic Biases in Reporting on Israel and Gaza

A new computational social science study published on arXiv reveals systematic biases in how major Western media outlets report on victims of the Gaza war. Researchers from NYU Abu Dhabi and the University of Padova analyzed over 14,000 articles from The New York Times, BBC, CNN, and Al Jazeera English using natural language processing and statistical methods. They found that Western outlets consistently portrayed Israeli victims as identifiable individuals with names and stories, while Palestinian victims were overwhelmingly depicted as anonymous collectives. This "identifiable victim" bias was just one of three patterns that created a distorted narrative framework for global audiences.

The study also documented "equalization bias," where Western media repeatedly referenced the October 7 attacks to create moral equivalence despite vastly asymmetric casualty counts, and "one-sided doubt casting" that selectively questioned Palestinian casualty figures. These patterns were largely absent in Al Jazeera English's coverage. The research demonstrates how AI-powered content analysis can move beyond anecdotal evidence to systematically quantify media framing at scale, with implications for media literacy, algorithmic content moderation, and understanding how information ecosystems shape public perception of complex conflicts. The 34-page main manuscript with 81 pages of supplementary material represents one of the most comprehensive computational analyses of war reporting to date.