Recent Lawsuits Set Precedent in AI Copyright Law, More Cases Pending

AI copyright law is set for major developments in 2025, beginning with a landmark decision on February 11 by a Delaware federal court in Thomson Reuters Enterprise Centre GMBH v. ROSS Intelligence Inc., writes Emilio B. Nicolas in ITPro.

The court ruled that Ross’s use of Westlaw’s headnotes to train its AI-driven legal research tool constituted copyright infringement. While not generative, Ross’s AI still functioned as a market substitute for Westlaw, weighing against a fair use defense.

This ruling signals a strict approach to AI training using copyrighted material, setting a precedent for ongoing legal battles. However, the decision’s broader implications remain uncertain. It could be overturned on appeal, and its reasoning may not apply universally—especially to generative AI systems.

Other pending cases, like Concord Music Group, Inc. v. Anthropic PBC, further complicate the landscape by tackling whether using copyrighted materials for AI training qualifies as fair use.

The case involves Anthropic’s AI assistant, Claude, highlighting a key unresolved issue: whether intermediate copies of copyrighted works made solely for AI training are legally permissible.

Courts are also grappling with whether AI-generated outputs that contain copyrighted material are infringing or transformative. Beyond training issues, AI’s ability to generate creative works raises deeper policy concerns about copyright ownership.

In Thaler v. Perlmutter, a court upheld the denial of copyright protection for AI-generated art due to the lack of human authorship—a decision now under appeal.

A similar case, Allen v. Perlmutter, argues for copyright protection when AI-generated art involves more human input. The US Copyright Office recently affirmed that AI-generated content is not eligible for copyright unless it includes substantial human authorship.

As these cases unfold, 2025 is poised to provide much-needed clarity on AI copyright law.