TL;DR:
The Ninth Circuit held on September 16, 2026 that DMCA Section 1202(b) does not reach generative AI outputs that omit copyright management information from training materials, effectively rejecting the “output theory” that AI outputs are “stripped copies.” This narrows the scope of potential 1202(b) claims against AI developers and impacts how litigants approach discovery, privilege, and data handling in AI-related disputes. For trial teams, the decision signals a shift in how to structure DMCA-based claims, design discovery protocols, and plan cross-examination around AI-produced materials. See Doe v. GitHub, Inc., No. 24-7700 (9th Cir. Sept. 16, 2026). (law.justia.com)
What happened and why it matters
Doe v. GitHub, Inc. presented two theories under DMCA Section 1202(b): an “input theory” that alleges removal or alteration of copyright management information (CMI) during the training or ingestion of code, and an “output theory” that contends the AI’s outputs themselves reproduce code without CMI, thereby violating the statute. The Ninth Circuit’s September 16, 2026 decision confirms that, at least in the context of generative AI outputs, the statute does not reach outputs that omit CMI from training works simply because the output is derived from those works. The court thus treats such outputs as non-violations of § 1202(b)’s “removal or alteration” provision for CMI, clarifying limits of DMCA liability for AI-generated results. The court’s opinion explicitly framed the ruling as addressing the “output theory” and did not resolve whether AI outputs could ever implicate other DMCA theories or civil liabilities. For practitioners, this narrows a frequently deployed route for policing AI-generated content and redirects focus to other potential claims or defenses. See Justia’s summary of the Ninth Circuit’s opinion and analysis, including the court’s treatment of the “output theory.” (law.justia.com)
The decision follows district court conclusions in the Northern District of California that had dismissed the DMCA claims at an early stage, with the Ninth Circuit affirming the limited scope of 1202(b) in the AI-output context. Legal observers and practitioners have highlighted that the ruling curtails a class of DMCA arguments centered on AI outputs that exclude CMI, which has been a focal point in major AI and software discovery disputes. See additional commentary from law firms analyzing the decision and its practical implications for tech litigation teams. (haynesboone.com)
Practical impact for trial teams
- Narrowed DMCA leverage: The decision directly affects how plaintiffs frame DMCA claims against AI providers. If a case hinges on an AI output’s omission of CMI, the Doe v. GitHub ruling suggests that such arguments may not proceed under § 1202(b) as to outputs alone. Litigation teams should reassess the viability of DMCA 1202(b) counts tied to AI outputs and consider alternative avenues, such as copyright infringement, contract, or trade secret theories, when pursuing or defending AI-related disputes. See the Ninth Circuit’s ruling and subsequent legal analysis. (law.justia.com)
- Discovery planning and prompts: While the ruling narrows DMCA 1202(b) exposure for outputs, it does not eliminate the broader importance of discovery around AI training data, prompts, and model behavior. Counsel should still anticipate requests for prompts, system prompts, and prompt-engineering notes where relevant to underlying claims, while carefully distinguishing discovery that implicates other protected information. In high-stakes AI matters, structured discovery protocols and audit trails for prompts remain prudent, even when 1202(b) claims are less likely to succeed. See industry and practitioner commentary on AI discovery practices in the wake of the decision. (ropesgray.com)
- Case strategy and pleading choices: Defense teams can leverage the ruling to argue that certain DMCA 1202(b) theories are unlikely to survive at the motion-to-dismiss stage or on summary judgment where outputs are involved. Plaintiffs, in contrast, may focus 1202(b) arguments on direct manipulation of CMI in training data or on other statutory or common-law theories, depending on the factual record. This separation of theories can shape early case strategy, preservation plans, and expert engagement. (haynesboone.com)
- Evidence handling and cross-examination: The decision underscores the need for careful interpretation of AI-generated materials during trial. While the 1202(b) theory may not support liability for outputs, cross-examination and fact-finding around how an AI system generated a particular result, including prompts used and data fed into the model, remain central to many AI disputes. Trial teams should build examination plans that explore the inputs and processes without conflating them with unauthorized handling of CMI under 1202(b). (ropesgray.com)
Implications for discovery strategy and defense readiness
- Prompt-level discovery remains important for other claims: Even with § 1202(b) narrowed, prompts, prompts logs, and model inputs can still be relevant to other theories, such as copyright infringement or misappropriation claims, or for understanding whether outputs infringed third-party rights. Counsel should tailor requests to avoid overbroad demands and to focus on information that directly supports the theory being pursued or defended. The Ninth Circuit’s narrowing of the 1202(b) path makes precise, theory-aligned discovery more critical than ever. (law.justia.com)
- Privilege and work-product concerns in AI litigation: As with other high-technology disputes, the production of AI-related materials implicates privilege and the work-product doctrine. Teams should preemptively design privilege logs and clawback procedures for training data, internal prompts, and engineering notes, ensuring that material privileged in nature remains protected while responsive information is produced in a defensible, auditable manner. Industry commentary on AI discovery protocols remains a useful resource for structuring these protocols. (ropesgray.com)
Practical next steps for litigators
- Reevaluate DMCA 1202(b) exposure in current AI cases: Review pending AI matters to determine whether 1202(b) claims are likely to survive given the Doe v. GitHub ruling, and adjust pleadings and motions accordingly. Where 1202(b) has been a central pillar, prepare alternative theories or narrow the claim focused on input data or other violations.
- Strengthen AI discovery plans: Develop robust but targeted discovery plans for prompts, training data provenance, and model documentation. Build a clear sampling and audit framework to demonstrate compliance with any court-ordered discovery protocols or protective orders, aligning with the court’s expectations for structured production.
- Integrate training for trial teams: Utilize practical training to simulate AI-related discovery and cross-examination scenarios. Objection Academy can help practicing teams rehearse prompt-logging, cross-examination on AI inputs and outputs, and the application of evidence rules to AI-produced materials, ensuring readiness for courtroom challenges. This kind of training aligns with the practical realities of AI litigation in the wake of Doe v. GitHub. (objectionacademy.com)
Takeaways for trial attorneys
- The Ninth Circuit’s stance on 1202(b) in the AI-output context narrows a common liability route in AI disputes, compelling litigants to pursue other theories or to tightly tailor their 1202(b) claims to train data manipulations and similar alternatives. Practitioners should recalibrate case strategy, focusing discovery and trial readiness on the broader evidentiary landscape surrounding AI systems and their outputs.
- For litigation teams, the ruling reinforces the value of methodical, auditable discovery protocols and a disciplined approach to AI materials in court. Training and practice for objections, evidence application, and trial-readiness remain essential, and Objection Academy provides a practical platform to sharpen these skills in the context of AI discovery and related evidentiary issues. (law.justia.com)
Sources:
- Doe v. GitHub, Inc., No. 24-7700 (9th Cir. Sept. 16, 2026). Justia summary of the Ninth Circuit opinion. (law.justia.com)
- Ninth Circuit decision discussed and analyzed by legal practitioners, including implications for DMCA 1202 claims in AI contexts. (law.justia.com)
- Commentary on AI discovery protocols and practical training considerations for trial teams, including Objection Academy. (ropesgray.com)