The text, essential for the survival of the media and cultural businesses, will have remained a dead letter. The discussion on the bill from MP Laure Darcos on the presumption of use of cultural content by artificial intelligence suppliers could not be held this Thursday, in the final hours of the parliamentary niche of the Democratic and Republican Left (GDR) group in the National Assembly. A disappointment for press publishers, a victory for the major American platforms, such as OpenAI, Anthropic or Google.
Approved by the Senate on April 8, the bill aimed to reverse the burden of proof when AI giants use information and cultural content to train their models. News media and cultural companies believe that AI actors are violating the fundamental principle of copyright by using content to which they have no rights.
« Incompréhensible »
The examination of the text could not be held, under pressure from the government and several deputies from the Macronist group EPR (Together for the Republic), including Guillaume Kasbarian and Prisca Thévenot. Taking refuge behind the risk of litigation within the European framework, they thus relayed the interests of the large American platforms. “It is incomprehensible to see the Assembly receptive to this speech when the text was voted on unanimously in the Senate,†Marc Feuillée, president of the Press Alliance, commented on Tuesday general information director and general director of the Le Figaro group.
“First serve human progress”
This failure provoked a reaction from the MoDem deputy for Finistère, Erwan Balanant, who passed a bill in March requiring digital platforms to pay financial compensation to publishers for the use of their content. “It’s a shame… We need sovereign and ethical AI. It cannot be done without fighting against the predation of these actors. Those who contributed to this obstruction think that technology will solve everything. I believe that innovation must first serve human progress.”