The U.S. government has filed a brief supporting OpenAI's position in cases concerning whether AI companies may use copyrighted material to train large language models, according to court filings reported by TechCrunch.
What Happened
The federal government submitted a filing in ongoing litigation that takes a permissive stance toward how AI companies source training data. The position aligns with OpenAI's argument that training AI systems on copyrighted works does not constitute infringement under existing law. The brief was filed as multiple copyright lawsuits against AI developers continue to work through the courts, with creators and publishers arguing their work was used without permission or compensation.
Why It Matters
The government's stance could shape how courts interpret fair use in the context of AI training—a decision that will affect the entire industry. If upheld, it would give AI companies broader latitude to build commercial products using data scraped from across the internet. Conversely, a ruling against the practice could force labs to license content or rebuild datasets, potentially slowing development and raising costs. For developers and enterprises building on AI systems, the outcome may determine whether foundational model capabilities remain widely accessible or become subject to stricter content licensing regimes.
The Bottom Line
The federal filing represents a significant intervention in high-stakes copyright litigation that could set precedent for how generative AI is developed and monetized. Courts have not yet ruled on the merits of these cases, and legal experts caution that regulatory positions do not guarantee judicial outcomes.