# Ninth Circuit Rejects DMCA Expansion in AI Training Case, Upholding Innovation In a landmark decision, the U.S. Court of Appeals for the Ninth Circuit has sided with internet users and programmers, rejecting an attempt to stretch a narrow provision of the **Digital Millennium Copyright Act (DMCA)** into a new source of copyright liability. This ruling carries significant implications for artificial intelligence development and broader digital creativity. ## The Core of the Dispute: DMCA Section 1202 The case revolved around Section 1202 of the **DMCA**, which specifically prohibits the intentional removal of copyright management information (CMI) – such as an author's name or a copyright notice – from a copyrighted work. The lawsuit was brought by a group of anonymous **GitHub** contributors, who alleged that **OpenAI** and **Microsoft** used their code, alongside billions of other works, as training data for their large language models (LLMs). The plaintiffs claimed that the new code generated by these LLMs was similar to their original works but lacked the CMI, thus violating Section 1202. ## A Crucial Distinction: Removal vs. Absence The Ninth Circuit, aligning with arguments put forth by organizations like the **Electronic Frontier Foundation (EFF)**, correctly distinguished between actively removing CMI from an existing copyrighted work and creating a new work that simply did not include CMI in the first place. The court emphasized that Section 1202 was intended to act as a safeguard for traditional copyrights in the digital realm, not to establish a novel, expansive right that could hinder otherwise non-infringing uses. ## Preventing a Floodgate of Liability Accepting the plaintiffs' theory would have opened the door to a new wave of liability for activities currently considered lawful. This could have severely undermined creativity and innovation across various fields, extending far beyond the specific context of AI development. The **EFF** highlighted that such an expansion could lead to costly lawsuits against legitimate users, including artists creating remixes, teachers adapting materials for classrooms, engineers reverse-engineering code for understanding, and search engines indexing the web. Independent software developers and smaller creators would have been disproportionately affected. While large corporations possess the resources to litigate such claims for years, individual programmers facing massive statutory damages might be compelled to settle, even if their underlying use was entirely lawful. ## Copyright Protection Remains Intact It is crucial to note that this ruling does not diminish existing copyright protections for programmers. They retain the ability to file copyright infringement claims if their work is unlawfully copied or reproduced by an AI model. Furthermore, the plaintiffs' contract claims against the AI companies are still pending. The Ninth Circuit's specific holding was narrow yet vital: the mere absence of copyright information from a *new* work does not, by itself, indicate that someone illegally *removed* it from an *existing* work. ## Navigating New Technologies with Established Law This decision underscores the judiciary's role in applying existing legal frameworks to emerging technologies without inventing new rights that could inadvertently stifle expression and lawful use. As new technologies continue to raise complex questions about copyright, courts must meticulously apply the rights authorized by Congress, ensuring a balance between protecting creators and fostering innovation. ## Additional Reading: * [9th Circuit Opinion in Doe v. GitHub](https://www.eff.org/files/2026/09/16/doe_v_github.pdf) * [EFF’s Amicus Brief in Doe v. GitHub](https://www.eff.org/files/2025/07/18/077.1_eff_and_public_knowledge_amicus.pdf)