The Extended Brief

Victory! Appeals Court Rejects Expansive New Copyright Claim

Brief by The AI News AI newsroom · Sep 18, 2026, 7:31 AM EDT edition

Original reporting by EFF Deeplinks — Joe Mullin · published Sep 16, 2026, 6:47 PM EDT

AI developers and other creators dodge a new copyright liability theory that could have made generating code without attribution metadata independently actionable.

Key points

• The Ninth Circuit rejected an attempt to expand DMCA Section 1202 into a new source of copyright liability. source ↗
• Anonymous GitHub contributors sued OpenAI and Microsoft, alleging LLM-generated code resembled theirs with copyright management information stripped. source ↗
• The court held that creating a new work without CMI differs fundamentally from removing CMI from an existing copyrighted work. source ↗
• The EFF argued the plaintiffs' theory would have exposed remix artists, teachers, reverse engineers, and search engines to costly lawsuits. source ↗
• The EFF said statutory-damages risk would have fallen hardest on independent developers, who may be forced to settle rather than litigate for years. source ↗

Practical applications

• Legal teams at AI companies can reassess Section 1202 exposure for training pipelines, while noting the ruling binds only the Ninth Circuit.
• Open-source developers pursuing claims over unattributed AI-generated code should treat traditional copyright infringement, not the DMCA CMI provision, as the viable route.
• Builders of tools that strip attribution from existing works should not read this ruling as cover — the court distinguished removing CMI from generating new works that never had it.

Context

The DMCA is a 1998 US copyright law, and its Section 1202 prohibits intentionally removing copyright management information such as an author's name or copyright notice from a work. This case centered on OpenAI and Microsoft's use of GitHub code as training data for their language models, and whether output resembling that code without its original notices violates the provision.

What to watch

• Whether the plaintiffs seek Supreme Court review or file in another circuit willing to adopt the broader Section 1202 theory.
• How courts apply the same reasoning in the other pending lawsuits over AI training on copyrighted code.

Related briefs

• OpenAI and Anthropic Revenues Dwarf Those of Chinese AI Models
• Everyone Says Datacenter Moratoriums Are Killing the US Buildout. We disagree
• Have it both ways: stay discoverable in search while disallowing AI training
• Houthis used Claude Code to develop missile guidance software: Anthropic

This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.