This paper explores how AI-driven storytelling can transform news articles into fictional narratives using structured retelling techniques. We introduce NewsReteller, a system that explores the generative capabilities of Large Language Models to create stories from news content through three distinct approaches: genre-based storytelling, which adapts narratives to established literary styles; structured storytelling, which reshapes events using predefined biased schemes (story skeletons); and data-driven storytelling, which emphasizes factual clarity and analytical framing. To assess the system’s ability to reinterpret factual content, we generated multiple stories from a single news article using each of these approaches. The results illustrate how different retelling strategies influence narrative framing, thematic emphasis, and information presentation, highlighting the potential of our method to generate creative reinterpretations of real-world events.
