Doe v. GitHub (Copilot)

Doe 1 v. GitHub, Inc., No. 4:22-cv-06823 (N.D. Cal., filed Nov. 3, 2022; Judge Jon S. Tigar), appeal decided by the U.S. COURT OF APPEALS FOR THE NINTH CIRCUIT in SEPTEMBER 2026. PROCEDURAL SPINE: the original DMCA § 1202(b) claim was dismissed 2024-07-05 for lack of an identical-copy allegation under the Ninth Circuit's identical-copy rule; surviving claims related to breach of open-source licence terms. THE NINTH CIRCUIT HAS NOW AFFIRMED, REJECTING THE PLAINTIFFS' EXPANDED § 1202 THEORY. HOLDING AS CHARACTERISED BY EFF (which filed a brief in the case): 'the absence of copyright information from a new work does not mean, by itself, that someone illegally removed it' — i.e. § 1202 liability requires removal of CMI FROM AN EXISTING COPYRIGHTED WORK, and is not established by the mere fact that model output lacks CMI that the training data carried. EFF's framing of the provision's purpose: § 1202 'was intended to serve as a backstop for traditional copyrights in the digital age — not to create a new, more expansive right to inhibit otherwise non-infringing uses.' WHY THIS IS INDEXED AS A PROTECTION DESPITE BEING A PLAINTIFF LOSS: the counterfactual liability would have run against USERS AND SMALL CREATORS, not only AI developers. On EFF's account, accepting the Does' theory would have exposed 'artists making remixes based on older works, teachers adapting works for a classroom presentation, engineers reverse engineering code' to costly federal suits, with the risk falling 'especially hard on independent software developers and other small creators' — large companies can litigate for years, whereas 'an independent programmer facing massive statutory damages may simply have to settle, even when their underlying use is completely lawful.' The decision therefore forecloses a statutory-damages lever against lawful reuse. WHAT SURVIVES, EXPRESSLY: (1) TRADITIONAL COPYRIGHT INFRINGEMENT CLAIMS ARE UNAFFECTED — 'They can still bring copyright infringement claims if someone uses a model to reproduce their code'; (2) THE PLAINTIFFS' CONTRACT CLAIMS AGAINST THE AI COMPANIES ARE STILL IN PLAY, i.e. the open-source licence-breach theory continues. RELATIONSHIP TO THE AI TRAINING-DATA LITIGATION GENERALLY: this narrows § 1202 as a route around the fair-use question rather than resolving fair use, which is untouched. LIMITS: the opinion itself was not retrieved, and NO CASE NUMBER, PANEL, DATE OR CITATION FOR THE NINTH CIRCUIT DECISION was obtained — see notes.