Rewriting Copyright Law for Generative AI Would Hurt Everyone

Why it matters

EFF has filed an amicus brief in In re OpenAI, one of several lawsuits in which publishers and creators accuse large AI companies of copyright infringement for training generative models on their works. EFF argues that the case reaches far beyond the parties: a ruling narrowing fair use would affect researchers, journalists, activists, artists, and people with disabilities who rely on flexible copying rules. The brief contends that early fears that AI models would closely reproduce training data have not materialized, and that some courts have already treated AI training as fair use because it serves a different purpose. EFF warns that requiring licenses for every work used would favor the largest technology and media companies, since only they could afford the costs, while creators would gain little. The organization stresses that copyright exists to promote public creativity and innovation, not to shield incumbents from competition, and that weakening fair use would jeopardize beneficial work like medical research, investigative journalism, and accessibility tools.