Michael Chabon and four other authors claim Meta and Open AI used their published works to train large language AI models.
