NAV AI
Logo

We live in an increasingly polarized world, where public discourse is shaped—and often divided—by differing political ideologies. This polarization is reflected not only in political debates but also in the news media we consume. Newspapers, while aiming to inform, often present information through lenses colored by political leanings. Understanding these underlying biases is essential for critically engaging with the news.

NAV AI (Navigator of Newspapers Political Orientation with Artificial Intelligence) offers a new way to navigate this landscape. It enables users to compare how leading open-source Large Language Models (LLMs) interpret the political orientation of newspaper articles. By submitting the URL of an article, users receive an analysis of the article’s ideological stance according to the Eysenck political compass, which spans from left to right and authoritarian to libertarian (learn more here: https://en.wikipedia.org/wiki/Political_spectrum). This tool helps shed light on how political perspectives are embedded in journalism, encouraging more informed and reflective media consumption.

Illustration

About the project

NAV AI was developed by the AI Readiness and Assessment (AIRA) research group, with the help of the University of Trento. Relevant scientific literature on this project:

Large Language Models' Detection of Political Orientation in Newspapers. Alessio Buscemi and Daniele Proverbio, arXiv preprint arXiv:2406.00018 (2024).

For further information, feel free to contact us: