In a Northern District of California case, Judge Vince Chhabria granted Meta partial summary judgment, ruling that Meta's use of copyrighted books to train its large language models constituted fair use under copyright law. The court found that training AI models is "highly transformative" and rejected arguments about market harm, noting that Meta's model could reproduce only about 50 words from any plaintiff's book. This ruling establishes precedent that training use for AI is categorically distinct from the original work's purpose and reduces training-data liability exposure for vendors deploying existing models on copyrighted material. For financial institutions, this narrows the probability and magnitude of balance-sheet provisions or revenue clawback risk tied to training-data IP litigation against major AI labs.
251 other tracked developments touching the same claims.