Appeals Court Rejects Attempt to Expand Copyright in AI and GitHub Case

The U.S. Court of Appeals for the Ninth Circuit has issued a ruling that the Electronic Frontier Foundation calls a major victory for internet users and programmers. At issue was an attempt to stretch a narrow provision of the Digital Millennium Copyright Act (DMCA) into a sweeping new basis for lawsuits.

The heart of the dispute

At the center of the case is Section 1202 of the DMCA, which prohibits intentionally removing from a protected work so-called copyright management information (CMI): the author's name, the copyright notice and similar details. OpenAI and Microsoft used code from GitHub — alongside billions of other works — to train their large language models. A group of anonymous GitHub contributors filed suit, arguing that the code these models output resembles their own but has been stripped of its authorship information.

The Ninth Circuit agreed with the arguments EFF laid out in its brief: removing authorship information from an existing work is fundamentally different from creating a new work that never contained such information in the first place. Section 1202 was conceived as a prop for traditional copyright in the digital age, not as a new, broader right to block perfectly lawful uses.

Why it matters

In EFF's view, adopting the plaintiffs' logic would have created an entirely new basis for liability over completely legal conduct. It would have put in the crosshairs artists who remix older works, teachers who adapt materials for a lesson, engineers who study others' code through reverse engineering, and search engines. Hardest hit would have been independent developers and small-scale authors: large companies can litigate in federal court for years, while a lone programmer facing enormous statutory damages is often forced to settle — even when their conduct is entirely lawful.

The court stressed the narrow but significant nature of its conclusion: the mere absence of authorship information in a new work does not by itself mean that someone unlawfully removed it. At the same time, copyright continues to protect programmers: if someone uses a model to reproduce their code, they can still bring an infringement claim. The plaintiffs' contract-based claims against the AI companies also remain in force.

EFF sums up: new technologies will continue to pose hard copyright questions for the courts, but they should be answered by applying the rights actually enacted by lawmakers, not by inventing new ones.

Source: EFF Deeplinks