Federal Judge Approves Anthropic's $1.5 Billion Author Settlement

Settlement Receives Final Judicial Sign-Off

According to The Verge AI, a federal judge has formally approved Anthropic’s $1.5 billion settlement with authors who alleged the AI company trained its models on copyrighted books without authorization or compensation. Judge Araceli Martínez-Olguín issued the approval order on July 21, finding the settlement would provide “meaningful relief” to the plaintiff class.

The resolution represents a landmark outcome in AI copyright disputes: plaintiffs’ legal counsel characterizes the $1.5 billion payout as the largest known copyright recovery in history. Each affected author or publisher receives roughly $3,000 per book deemed to have entered Anthropic’s training pipeline.

Widespread Claim Participation and Lingering Disputes

Anthropic general counsel Aparna Sridhar confirmed to Reuters that more than 91% of authors and publishers eligible under the settlement terms have already filed claims for their allocated shares. This high participation rate suggests relatively broad acceptance of the per-book valuation across the author community.

However, the settlement does not represent universal closure on Anthropic’s copyright exposure. The company continues to face separate lawsuits from parties including Chicken Soup for the Soul and additional authors who contend that $3,000 per book undercompensates for the commercial value extracted from their intellectual property during model training.

Legal Path to Approval

The lawsuit originated in 2024 when authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson filed suit against Anthropic. An initial ruling by the now-retired Judge William Alsup granted Anthropic partial victory on some claims while allowing a class-action case to proceed based on allegations of “Napster-style downloading of millions of works.” Anthropic agreed to the settlement framework in September 2025, receiving preliminary approval from Alsup before the current final judgment.

Why This Matters

This settlement establishes a precedent for monetary remedies in AI copyright disputes, setting a floor valuation ($3,000 per book) that competing interests—both authors seeking higher rates and AI companies seeking lower ones—will reference in future negotiations. For authors and publishers, it demonstrates that courts will enforce copyright protections against large language model developers, though ongoing litigation from dissenting claimants suggests the per-book figure may face challenge. For Anthropic and the broader LLM industry, the approval signals that using copyrighted training data without explicit licensing creates substantial financial and reputational liability, likely influencing future training data acquisition strategies across the sector.