Overview

• Judge William Alsup ruled on June 23 that Anthropic can continue training on scanned books from legally obtained copies provided the originals are destroyed, dismissing most infringement claims except those related to its initial pirated library.
• On June 25 Judge Vince Chhabria granted summary judgment to Meta, concluding that plaintiffs failed to show substantial market harm but stressing the decision does not legalize broad use of copyrighted material for AI training.
• Both companies remain entangled in separate piracy allegations, with Anthropic facing a forthcoming trial and Meta required to negotiate remedies with its plaintiffs.
• More than 40 similar copyright suits against leading AI developers are pending, and their outcomes will determine whether licensing frameworks are needed or firms can rely on fair use defenses.
• Legal experts and creators warn that the lack of clear copyright standards risks undermining authors’ livelihoods or driving AI training practices into legal gray areas.