The Le Monde group adopts a charter on artificial intelligence

The Le Monde group adopts a charter on artificial intelligence

The media landscape has experienced two revolutions in recent decades: the advent of the Internet in the 1990s and then the rise of social networks in the 2000s. The so-called generative artificial intelligence (AI) (capable of creating texts, images or videos) will it in turn disrupt a decidedly unstable ecosystem? The question has come up repeatedly in numerous debates, conferences, seminars and roundtables organized since the launch of a general public version of the ChatGPT conversation robot in late 2022.

Without further ado, the Le Monde group has decided to adopt a charter on the use of AI by its editorial staff, a text intended to complement its charter of ethics and professional conduct. The document was adopted by the Le Monde group’s ethics and professional conduct committee on December 15, 2023. The journalistic companies of the various titles involved (International mail, Le Monde, Telerama And Life) were involved in the discussions. This text should be legally adopted by the next general meeting of the Société Éditatrice du Monde.

The document consists of four articles and establishes in the preamble a principle that has the merit of clarity: “Under no circumstances can AI replace humans in our journalistic productions and performance. » This charter is addressed both to the journalists of the press group and to those who read it. “It is essential to reaffirm to our readers the principles that govern us to maintain the trust they place in the titles of the Le Monde group”so it is claimed.

Article 1 guarantees readers that no editorial content can be designed outside the communities of human intelligence that constitute the editorial staff of the Le Monde group. “Generative artificial intelligence (…) cannot in any way replace the editorshe specifies. The use of generative AI is only permitted under strictly regulated conditions as an aid to editorial production. »

The charter on the use of AI within the Le Monde group requires that any other use of generative AI be systematically and explicitly reported to the reading public – “The use of generative AI to create images is prohibited”, it has also been stated. One of the conditions for using generative AI is that it allows this “to improve the quality of information content”.

Treasures of ingenuity

The titles of the Le Monde group do not prevent themselves from using certain tools that use artificial intelligence. The world in English is an example: this English version of the site World has been offering several dozen articles every day since April 2022, produced by the editorial staff and translated into English. Each of these articles is translated in three stages: the first “translation layer” is performed by the automatic translation software DeepL, the second by a professional translator (human!), the third by an English-speaking journalist.

This project would not have been possible without the improvement in the performance of the translation software: the cost in labor hours of a 100% human translation would have been daunting given the expected revenues. It is the advancement of AI that has made it possible to believe that it could find its economic equilibrium within a reasonable time. Ultimately, several jobs for English-speaking journalists were created within the newsroom WorldAI is therefore not necessarily synonymous with job destruction.

Another example of an additional service offered by The world to its readers thanks to controlled use of AI: offering an audio version of the articles. These ‘listening articles’ are automatically generated based on the written version of the text and the recorded voices of three actors and three actresses. Even if imperfections remain in the pronunciation of certain words and acronyms, they will be gradually erased by the AI. Users of the application World now have access to this audio version for all articles.

The AI ​​finally helps the team that takes care of the moderation of the comments of the subscribers of the World to identify posts that do not respect our contribution charter. Here again, nothing can replace a human eye and a human brain to thwart the treasures of ingenuity that some people display by flirting with the limits set by the set rules.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *