No items found.
Success story
Spinoza: A generative artificial intelligence tool created with and for journalists, dedicated to augmented and ethical journalism
Launched by Reporters Sans Frontières (RSF) – Reporters without borders – and the Alliance de la presse d’information générale with the technological support of Ekimetrics, the Spinoza project offers a prototype generative AI tool designed in close collaboration with journalists to respond to the challenges of their profession. This experiment lays the groundwork for a responsible use of AI in newsrooms dedicated to news.
120+ media mobilized
The Spinoza project has gathered 120 publications from 12 French press companies to co-construct the tool with journalists and publishers.
6 specialized databases
Scientific, legal, public, environmental, journalistic, and AFP data: A robust documentation architecture, designed to guarantee the reliability of answers.
93% of journalists interested
According to a study by RSF, 93 % of the journalists surveyed plan to use generative AI in their professional practice.
Discover
What we did
01
Challenge
- A need for a reliable and transparent tool to analyze large volumes of data (scientific reports, legal documents, articles) on technical subjects
- Test AI in a controlled environment that respects journalistic ethics
- Produce a tool that cites sources and traces information
02
Our approach
- Collaborative design with 120 editorial teams brought together by the Alliance to define usage, limits, and expectations
- Development of a prototype by Ekimetrics based on selected databases, an embedding engine, and calibrated prompts
- Launch of a pilot phase on the theme of climate change, with journalist testers
03
Outcome
- An editorialized generative AI, designed as a research and verification assistant: It doesn’t write in the journalist’s place but helps them to structure their documentation using verified and traceable sources.
- Reliable summaries of complex subjects (such as climate change), resulting from the complementarity between scientific, regulatory, and journalistic bases, with direct links to original documents.
- A transparent, ethical, and customizable tool, already used by several journalists in France, whose open-source version (scheduled for 2025) will allow editorial teams to adapt it to their specific needs, while collectively discussing responsible use of AI in the media.
Challenge
Our approach
Outcome
Explore
Other Success Stories
Get in touch
Connect with our Data Science experts
Oops! Something went wrong while submitting the form.