Generative AI promises the world — but how do we really know if it delivers what we need in the newsroom? In this session, we go beyond the hype and show how to test and compare AI in ways that produce reliable answers, not just gut feelings.
We break down journalistic tasks into concrete criteria, set up practical evaluation frameworks, and demonstrate how to identify which models and prompts actually perform best. The goal: less guesswork, more control — and AI that genuinely helps you produce better journalism.
