In one of the most emblematic lawsuits between rights holders and AI developers, Meta has just won an important victory, but the judgment points out that with better arguments, the authors would probably have won. This week is definitely not a good one for rights holders. On June 23, a San Francisco judge accepted as valid the « fair use » exception in the Bartz case, pitting book authors against the generative AI company Anthropic. It’s a disappointment for the creative world, even if the judge warns that Anthropic will have to pay a hefty penalty for having used pirated works to build the library that trained its LLM Claude. Yesterday, another blow fell: Judge Vince Chhabria, also from San Francisco’s District Court, decided in favor of Meta in a similar case, in which several book authors felt that Meta had infringed their copyright by training LLaMa with their works, without authorization. We reported on this case, and reviewed Meta’s arguments in detail, here. Astonishingly, Judge Chhabria seems to adopt his decision reluctantly, and warns the AI developers that it’s because…
