Publishers Weekly: SCOTUS's Cox ruling may open a new inducement front in AI copyright fights

Publishers Weekly analyzes the Supreme Court's unanimous Cox ruling this week (which rejected a $1 billion copyright judgment against ISP Cox Communications on the ground that an ISP is liable only if its service was designed for illegal activity or actively induced infringement) and argues the reasoning may open a second front in publishers' and authors' AI copyright fights. Unlike an ISP's neutral conduit, large language models are, by design, built to generate prose, poetry and dialogue stylistically and substantively indistinguishable from human authors' work, the piece says; plaintiffs' attorneys may argue that training on copyrighted text while optimizing for human-quality creative output is a service 'tailored for' infringement or induces it, beyond the existing training-data-ingestion question. It cautions that the music industry's costly campaign against Cox yielded a fractured, narrowed result.