Shaping Media Narratives on the Palestinian Issue: A Comparative Study of Key Generative AI Tools

Abstract

This research examines the increasing influence of generative AI on media narratives surrounding the Palestinian issue. Through a comparative analysis of AI-generated media content from models like GPT-3.5, DeepSeek-V3, Mistral AI, and Jais, using targeted prompts, we investigate the technology's impact on news formulation, information dissemination, and potential for bias. Combining AI-generated text analysis with media content analysis, the study assesses accuracy, bias, and the capacity to shape public opinion. Preliminary findings indicate that generative AI can both diversify media narratives and, depending on training data and origin, reinforce existing biases, compromising objectivity and neutrality. The research offers recommendations for responsible AI application in media, encouraging journalists to utilize models that present factual information neutrally to ensure balanced reporting on complex political issues like the Palestinian issue.