OpenAI and GEDI partner for Italian news content
OpenAI
OpenAI just struck a deal with GEDI, one of Italy's biggest news publishers, to feed their journalism into ChatGPT. It's another sign AI companies are paying for credibility instead of just scraping it.
Based on reporting by OpenAI — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI has signed a strategic partnership with GEDI, the Italian media group behind titles like La Repubblica and La Stampa, to bring the publisher's news content directly into ChatGPT. The announcement is thin on dollar figures and technical specifics, but the direction is unmistakable: OpenAI wants licensed, local-language journalism inside its products, not just whatever it can pull from the open web.
Italy matters here for reasons beyond market size. GEDI's newsrooms carry weight in Italian public life, and Italy's data protection authority has been one of the more aggressive regulators in Europe when it comes to questioning how AI firms handle personal data and copyrighted material. A formal content deal with a respected domestic publisher reads as much like a regulatory hedge as a product upgrade.
This fits a pattern OpenAI has been building for over a year, following similar arrangements with the Associated Press, Axel Springer, Le Monde, and a handful of other outlets across different countries and languages. Each deal follows roughly the same logic: publishers get paid and get their content attributed inside a widely used chatbot, while OpenAI gets to say its answers on current events are sourced from real reporting rather than an unlicensed scrape job.
What's less clear is how much this actually changes for the average ChatGPT user in Italy or anywhere else. Licensing agreements don't automatically translate into visibly better answers, and the mechanics of how GEDI's articles will be surfaced, summarized, or credited inside conversations haven't been detailed. Still, for an industry facing lawsuits over training data on multiple continents, quietly signing up respected publishers one country at a time is a cheaper and less risky path than fighting every copyright claim in court.
My take — AI-written commentary, not fact-checked reporting
I write for an independent outlet with no licensing deal with anyone, so forgive me for noting that these partnerships are as much legal insurance as product features. Good on GEDI for getting paid, but let's not pretend this settles the bigger fight over what AI companies owe the people whose work trained them in the first place.
Read more about this at: OpenAI