TL;DR:
The Ninth Circuit on September 16, 2026 held that outputs produced by generative AI tools do not constitute “copies” that remove or alter copyright management information under DMCA section 1202(b). In Doe v. GitHub, Inc., the court affirmed dismissal of the DMCA claims against GitHub, Microsoft, and OpenAI, ruling that AI-generated outputs are newly created works and not copies that carry or lose CMI. The decision effectively narrows a potential DMCA-based path for plaintiffs alleging attribution or licensing violations in AI-generated code, and it leaves room for other theories such as copyright infringement or contract-based claims. Trial teams should rethink DMCA leverage in AI-forward cases and adjust discovery and evidentiary strategies accordingly. The ruling also underscores the value of structured objection and evidence-training practice for AI-related disputes, an area where tools like Objection Academy can help litigation teams prepare. The decision is a pivotal data point for practitioners navigating AI training data, prompts, and output authenticity in federal litigation.
What happened and why it matters
On September 16, 2026, the United States Court of Appeals for the Ninth Circuit issued its opinion in Doe v. GitHub, Inc., addressing whether DMCA section 1202(b) can be violated by AI-generated outputs that omit copyright management information (CMI) such as attribution or license terms. The panel held that AI outputs are not “copies” of existing works containing CMI, but rather newly generated works. Consequently, the court affirmed the district court’s dismissal of the DMCA claims under §1202(b). The decision also declined to consider an “input theory” of liability because the plaintiffs forfeited that theory in the district court. In short, generating outputs without CMI does not violate 1202(b) as a matter of law. This ruling reduces the potential DMCA leverage for plaintiffs in AI-related disputes and shapes how litigants frame claims around AI training data and generated content. The opinion is published in the Ninth Circuit’s official record as of September 16, 2026. (cases.justia.com)
The court’s reasoning centers on the distinction between copies that already exist and are subsequently altered, and outputs that are newly created by the AI system. The panel acknowledged the DMCA’s goal of protecting copyright management information, but found that 1202(b) targets the removal or alteration of CMI in existing copies rather than the generation of new works by an AI model. Because Copilot and Codex produce outputs that “never contained” CMI in the first place, they do not remove or alter CMI from a copy, and thus do not trigger liability under §1202(b). This is the governing interpretation of the statute in the Ninth Circuit. The court also stated that the plaintiffs had forfeited the input theory by not pursuing it in the district court, limiting the scope of the appeal. (cases.justia.com)
Practical implications for litigators
- Narrowed DMCA leverage in AI-forward cases: The ruling narrows a potential liability path for defendants whose AI-generated outputs may raise attribution or licensing concerns. Plaintiffs cannot rely on DMCA §1202(b) to claim liability for AI-generated outputs that lack CMI, at least in this circuit. Litigators should reassess whether a 1202(b) claim is viable in AI output disputes and plan alternative theories of liability, such as direct copyright infringement or breach of license terms. (cases.justia.com)
- Focus on alternative theories and evidence strategies: With 1202(b) less available, practitioners should build stronger claims under traditional copyright theory or contract-based theories, and emphasize the independent conduct of defendants, licensing terms, or representations around attribution in training data. Discovery strategies should align with those theories, including procurement of licensing documents, training data disclosures where feasible, and any contractually bound representations around AI outputs. (cases.justia.com)
- Implications for discovery and evidence handling: The decision spotlights the need for robust handling of AI inputs and outputs in discovery. Firms should consider how to structure requests for training data provenance, prompts used during generation, and the chain of custody for AI-produced materials when pursuing or defending AI-related claims. While the Ninth Circuit did not rely on training-data disclosures to support §1202(b) claims, practitioners should anticipate future disputes over the admissibility and authenticity of AI-generated outputs and the metadata attached to those outputs. (cases.justia.com)
- Cross-jurisdiction considerations: While Doe v. GitHub is a Ninth Circuit decision, other circuits may reach different conclusions about DMCA §1202(b) in the AI context. Counsel should monitor how neighboring circuits treat “output theory” versus “input theory” in DMCA claims and be prepared to tailor arguments by jurisdiction. Industry commentary from law firms indicates practitioners should not assume a nationwide DMCA remedy exists for AI-generated outputs, making multi-district litigation strategies more complex. (ropesgray.com)
How trial teams can translate this into practice
- Reframe pleadings around AI content: When drafting complaints or motions in AI-generated content cases, emphasize ownership, licensing, and originality of the outputs, rather than relying on DMCA §1202(b) to impose liability for AI outputs. Prepare to demonstrate that outputs are new works and that any liability would have to arise from other theories. (cases.justia.com)
- Build evidence-readiness for AI disputes: Create a litigation plan that documents the provenance of AI outputs, the prompts used, and the chain of custody for materials introduced at trial. This preparation can help address authenticity, reliability, and attribution issues, which are central to many AI-related disputes. Consider using trial-readiness tools and practice regimes to stress-test your objections and framing for AI evidence. Objection Academy offers practice resources that can help trial teams drill objections and evidentiary handling in AI-led disputes, complementing a juror-facing strategy. (objectionacademy.com)
- Prepare for jurisdictional variance: In parallel with the Ninth Circuit ruling, anticipate how other circuits may treat 1202(b) claims in similar AI contexts. Develop parallel pleadings or motions in case a case law trajectory shifts in other jurisdictions. Consulting recent firm analyses can help tailor strategy to circuit-specific nuances. (haynesboone.com)
Objection Academy and trial-readiness considerations
For trial teams aiming to stay ahead in AI-related disputes, integrating objection-focused practice can strengthen courtroom readiness. Objection Academy’s materials and analyses on AI discovery and evidence readiness can help litigators frame and defend objections around AI prompts, outputs, and metadata in real time. While the Ninth Circuit’s Doe v. GitHub decision narrows DMCA 1202(b) liability for AI outputs, effective courtroom advocacy remains essential to handle evidentiary issues, authentication, and credibility of AI-generated materials. (objectionacademy.com)
What to watch next
- Monitor whether other circuits adopt the Ninth Circuit’s reasoning or carve out different approaches to DMCA §1202(b) in AI generated content scenarios. Practitioners should track subsequent cases and official guidance from courts on AI-generated evidence, prompts, and outputs. The Doe v. GitHub ruling will likely influence lower court rulings pending similar disputes in other jurisdictions. (cases.justia.com)
- Stay aligned with evolving evidence rules and AI governance developments. The federal judiciary and bar bodies continue to issue guidance on AI in litigation, including how to tag, disclose, and manage AI-generated content in discovery and trial. (uscourts.gov)
Sources:
- Doe v. GitHub, Inc., No. 24-7700, 9th Cir. (Sept. 16, 2026). Download PDF: https://cases.justia.com/federal/appellate-courts/ca9/24-7700/24-7700-2026-09-16.pdf?ts=1789576246. The Ninth Circuit held that AI outputs are newly generated works not copies containing CMI under DMCA §1202(b), and that the input theory was forfeited. (cases.justia.com)
- Doe v. GitHub, Inc. — Justia case page summarizing the decision and the DMCA framework. (law.justia.com)
- Ninth Circuit commentary and practitioner analysis: Haynes Boone, Ninth Circuit Rejects DMCA Section 1202(b) Theory Against Copilot Output in Doe v. GitHub (Sept. 16, 2026). (haynesboone.com)
- Objection Academy coverage and relevance to AI-related evidence practice (trial readiness and objections): Objection Academy article examining the Ninth Circuit ruling and related AI-discovery themes. (objectionacademy.com)
Note: This analysis centers on a timely controlling decision within the last 45 days and translates it into actionable guidance for trial teams.