This study applies a rule-based Natural Language Processing (NLP) approach to analyze sentiment in 7,596 multilingual news headlines covering the 2025 India–Pakistan conflict. Headlines were translated to English, normalized, and classified using a lexicon-based model (TextBlob) into five categories: pro-India, pro-Pakistan, unclear-positive, unclear-negative, and neutral. Sentiment distribution was examined across countries and over time. Results show that 92.5% of headlines were neutral, with limited overt bias. Mixed sentiments were more common when both nations were mentioned, especially in Indian media. A temporal shift from neutral to sentiment-rich headlines between April and May was observed. While effective for preliminary analysis, the model’s limitations suggest the need for more sophisticated NLP tools to capture nuanced geopolitical sentiment.