Anthropic vs. Music Publishers: What the Fair Use Ruling Means for AI Training

Seed story: "Anthropic, music publishers square off on 'fair use' defense in AI copyright case" (Reuters) · search original Written from facts verified across 3 news report(s) — original explainer, not a copy or translation. Sources listed at the end.

With US District Judge Eumi Lee set to weigh the fair use defense in a high-stakes hearing, the outcome of this 2023 lawsuit against Anthropic could define whether AI companies must pay licensing fees for training on copyrighted music. The decision carries significant weight for creators, as it may establish a new liability framework that influences pending copyright cases against tech giants like OpenAI and Meta.

The Standoff in San Jose

On October 21, 2026, U.S. District Judge Eumi Lee presided over a critical hearing in San Jose regarding the copyright lawsuit against Anthropic. While the judge did not issue a ruling on the core fair use defense, she actively questioned both legal teams. This procedural step highlights the complexity of the dispute, which centers on whether training AI models on copyrighted lyrics constitutes infringement.

The hearing focused on the tension between transformative use and market harm. Key arguments included:

Anthropic’s claim that using lyrics is a transformative act protected by fair use.
Publishers’ assertion that the AI generates competing content that undermines human songwriters.
The scale of the issue, with nearly 120,000 requests for synthetic lyrics reported in six months.

For creators, this standoff is pivotal. The outcome will likely influence pending cases against other tech giants, directly impacting how rights are defined and compensated in the AI era.

The Core Legal Battle

At the heart of this dispute lies a fundamental clash over the definition of "transformative" use. Anthropic’s attorney Sonal Mehta contends that ingesting song lyrics to train Claude is a transformative act. She argues this process helps the model understand human language, ultimately enabling progress in science, business, and education. For creators, this stance suggests that training data usage might be shielded from liability if it serves a broader educational or technological purpose.

Conversely, the publishers paint a starkly different picture of market harm. Their attorney Matt Oppenheim asserts that Anthropic misuses lyrics to generate competing content that directly undermines human songwriters. To quantify this threat, they cited nearly 120,000 requests for synthetic lyrics over just six months.

Transformative Purpose: Anthropic claims training aids language comprehension and societal progress.
Market Competition: Publishers argue the AI produces direct substitutes for human work.
Scale of Output: Cited figures show massive volume of generated lyrics.
Artist Impact: The case involves works by Beyoncé, the Rolling Stones, and the Beach Boys.

This debate is critical for creators because it determines whether AI companies can legally bypass licensing agreements by claiming their output is sufficiently distinct from the original source material.

Split Rulings and Market Impact

The legal landscape for AI training remains fragmented due to conflicting precedents from Northern California. Judge William Alsup sided with Anthropic in 2025, supporting the fair use defense for model training. Conversely, Judge Vince Chhabria ruled in favor of Meta but issued warnings against overly broad fair use claims. These split decisions create significant uncertainty regarding liability for tech companies.

Alsup’s 2025 ruling favored Anthropic’s transformative use argument.
Chhabria’s Meta decision limited the scope of fair use protections.
Judge Lee’s October 2026 hearing highlighted these divergent interpretations.

This inconsistency complicates how publishers and developers assess risk. If Judge Lee aligns with Alsup, it may validate training on copyrighted lyrics. However, if she follows Chhabria’s caution, the defense could be narrower. Creators should monitor these rulings closely, as they directly impact whether their works can be used without compensation in future AI models.

Implications for Creator Contracts

Financial Shifts in the AI Era

If Judge Lee rejects the fair use defense, the legal landscape shifts dramatically. AI companies would likely be forced to negotiate licensing agreements, fundamentally altering how revenue flows to rights holders. This transition moves the industry from a "free-for-all" training model to a structured market where access to creative works carries a price tag.

For independent musicians and writers, this change could reshape their income streams:

Direct Licensing Fees: Publishers may pass through a portion of AI training costs to individual songwriters.
Collective Bargaining Power: Creators might gain leverage through collective agreements with AI firms.
Market Competition: As noted by publishers, AI-generated content could still overwhelm human works, potentially depressing traditional sales.

Ultimately, the ruling determines whether creators are compensated for their intellectual property or remain unprotected in the face of synthetic competition.

Strategic Moves for Independent Creators

As the legal landscape shifts, independent creators must proactively secure their intellectual property. With Judge Eumi Lee’s October 21, 2026, hearing leaving the fair use defense unresolved, uncertainty remains high. Creators should review existing contracts to ensure they retain full control over how their work is used in AI training.

Key protective steps include:

Audit Licensing Agreements: Verify that current deals do not implicitly grant AI training rights.
Monitor Market Trends: Track the nearly 120,000 synthetic lyric requests cited by publishers to understand potential market saturation.
Diversify Revenue: Reduce reliance on platforms that may face future restrictions or licensing mandates.

Given that this case could influence pending suits against OpenAI and Meta, staying informed is critical. Creators should prepare for potential licensing opportunities or new restrictions, ensuring their rights are explicitly defined before broader industry standards are established.

FAQ

What is the current status of the copyright lawsuit against Anthropic?

US District Judge Eumi Lee held a hearing on October 21, 2026, but did not issue a ruling at that time. The judge questioned both sides regarding their fair use defenses in the case brought by music publishers UMG, Concord, and ABKCO.

Why are music publishers suing Anthropic for using song lyrics?

The publishers allege that Anthropic infringed their copyrights by using their lyrics to train the AI model Claude. They claim this practice generates competing content that harms human songwriters, citing nearly 120,000 requests for synthetic lyrics over a six-month period.

How might this case affect other AI copyright lawsuits?

The outcome of this case could influence the results of other pending copyright lawsuits against tech giants like OpenAI and Meta Platforms. Previous rulings in Northern California have been split, with some judges siding with AI companies on fair use grounds while others have issued warnings against broad claims.