French publishing has a new AI deal. LVMH's media division, Groupe Les Echos-Le Parisien, has signed a strategic agreement with OpenAI that covers two things: AI tools for its own staff, and a place for its journalism inside ChatGPT's answers.
On the tools side, the group will roll out ChatGPT Enterprise across its newsrooms and business teams, with training tailored to each role. The suggested uses are document research, comparing documents and first-draft translation, with the stated aim of freeing journalists for investigation, analysis and reporting. The group also wants AI for personalisation, to keep readers and advertisers and grow revenue.
On the distribution side, ChatGPT will show summaries and excerpts from the division's titles when users ask relevant questions, naming the source and linking to the original article. The division is large: 17 brands, including Les Echos, Le Parisien, Challenges, Paris Match, L'Opinion, Investir and L'Agefi, plus Radio Classique and Mezzo. It has 1,800 employees, nearly 700 of them journalists, and reaches more than 27 million people in France each month. OpenAI's EMEA managing director, Emmanuel Marill, said the deal respects the titles' editorial independence. Neither side disclosed the money involved.
This fits a pattern that began with Axel Springer in late 2023 and continued with Le Monde and Spain's Prisa Media in 2024, both of which agreed to ChatGPT showing summaries with attribution and links. With Le Monde and now Les Echos-Le Parisien signed, OpenAI has two of France's major news groups. For readers who use ChatGPT for news, it means more answers grounded in named French sources rather than unattributed web text.
The doubts are familiar. Publishers have little public evidence that links in chatbot answers send meaningful traffic back, and a good summary can remove the reason to click. The announcement does not say whether the content will be used to train OpenAI's models, which earlier deals with other publishers included, or how long the agreement lasts. The group has not said how it will check AI-assisted research or translation before publication.