Despite the Emmanuel Macron’s great reluctance to acknowledge the relevance of this subject, the principle of respect for copyright by generative AIs was given lip service in the final declaration of the AI Action Summit, which nonetheless adds nothing new in relation to the Hiroshima Process. We had been following the saga of negotiations between culture and organizers in the run-up to the Summit for Action on Artificial Intelligence. The creative sector wanted more than a folding seat at the Summit, and demanded a commitment from the French state, and other participating states, to recognize their rights and help protect them from AI developers, particularly as regards works used to train LLMs and other generative tools. In the end, a round table did take place during the summit itself, but without the developers: we reported on it here. And, as we shall see, the summit’s final declarations mention culture, but go no further than the Hiroshima Process Code of Conduct, already agreed between 49 states as early as 2023. « We need to be in the race before we know how to regulate » For culture, the Summit had got off to a bad start with, on Sunday evening, a TV interview with Emmanuel Macron, who twice refused to talk about copyright and artists’ rights. Palki Sharma, an Indian journalistic celebrity,…
