Reuters flags 2026 as a decisive stretch for the copyright-and-AI docket, with federal courts set to rule on whether training generative models on copyrighted works qualifies as fair use. For data brokers, publishers, and AI labs alike, these rulings will shape licensing economics far beyond any single case—determining whether scraped or licensed data becomes the default compliance path.
Companies sitting on large training corpora should expect the outcomes to ripple into contract terms, indemnification clauses, and data-sourcing audits industry-wide.
AI copyright battles enter pivotal year as US courts weigh fair use
— Reuters
