Predicting international events like political upheavals or economic shifts is a complex puzzle. Researchers are turning to AI, hoping to build systems that can analyze news and forecast future global happenings. A new research paper introduces "WORLDREP," a dataset designed to train AI for this challenging task. Existing datasets often miss the nuances of international relations, simplifying complex multilateral interactions and often mislabeling the relationships between countries. WORLDREP aims to fix this by using the power of large language models (LLMs). These LLMs can analyze news articles, identify the countries involved, and even score the relationships between them, from full cooperation to outright conflict. To check the AI's work, experts in political science reviewed the LLM's analysis, confirming its accuracy. Initial experiments show that AI models trained with WORLDREP do a better job of predicting relationships between countries compared to models trained on older datasets. This suggests that AI could indeed have a future in forecasting international events, potentially aiding policymakers and analysts in navigating the complex world of geopolitics. However, challenges remain, including the potential for AI bias and the need for even more refined models that can capture the full complexity of global events.