Bodoni: AI tools for the news ecosystem

Context

Bodoni is an AI-native platform for journalists and newsrooms, developed by Geckosoft in collaboration with FIEG (the Italian federation of newspaper publishers) and LUISS Data Lab, and funded by Google.org.

Problem

Disinformation and AI-generated content are undermining the quality of information, and ultimately trust in journalism itself. Journalists needed tools to navigate an increasingly complex news landscape — surfacing patterns, verifying facts, drafting assistance — without those tools becoming one more layer of complexity on top of an already demanding job. The brief asked for a system that stayed genuinely useful for a newsroom, not just technically impressive.

What guided the design has been the concept “Make the AI show its work”. Every suggestion, cluster and CoEditor response had to be traceable back to a verified source — trust engineered into the interface, not asked for on faith.

The platform

We developed a modular editorial experience that combines analytical precision with generative power. Starting with a clear, functional design, we built a system that guides users through data, news, and sources using a semantic graph, smart filters, and verification tools. The interface integrates narrative and operational sections, placing a strong emphasis on readability and writing support via Co-Editor—an AI-powered assistant. The entire system is designed to evolve alongside the editorial team, maintaining a balance between technological complexity and journalistic accessibility. Bodoni is built by a mixed team of Geckosoft engineers working alongside LUISS Data Lab researchers; my role covered product definition and UX/UI across the product, in tandem with the great people of Google. I wrote about the UX research and the backstage of all the project on Geckosoft blog here. More info on the project here.

Explore: a graph to map the news landscape

The Explore feature of Bodoni provides an interactive graph that maps the relationships between news articles, topics, and sources. This visual tool allows users to explore how stories connect, revealing patterns and unique angles by highlighting outliers and anomalies. Filters enable customization by source, time period, location, and other parameters, helping journalists discover emerging trends and diverse perspectives.

News feed: clustering by topic

The News section clusters articles by topics, allowing for a structured comparison of coverage from various outlets. This organization helps journalists identify differing viewpoints, similarities, and contrasts in media narratives, enhancing their understanding and enabling a more comprehensive presentation of events.

CoEditor: the AI assistant built around trustworthiness

AI CoEditor assists journalists in writing by providing real-time suggestions, verifying information, and referencing reliable sources. It integrates a chatbot that answers current affairs questions using only verified facts, ensuring articles are accurate and helping reduce the spread of misinformation. This tool not only improves the writing process but also supports journalists in maintaining high standards of journalistic integrity.