A new AI pipeline, Data2Story, uses seven agents to turn CSV files into interactive news articles with verifiable evidence. The system, developed by Oxford and Stanford researchers, includes a ‘Inspector’ panel that tracks every claim and asset. The base model is Claude Opus 4.7, with OpenRouter models for images, video, and audio. The results show that 53 readers rate agent articles higher than human originals, highlighting the potential for AI to augment data journalism. This innovation matters because it could significantly reduce the time and effort required for data-driven investigations, enabling journalists to focus on higher-level analysis and storytelling.