Anthropic's $1.5B Copyright Settlement Sees Legal Fees Reduced

In a landmark copyright dispute within the AI industry, the debate over how to "divide the pie" has resurfaced. Recent court filings reveal that the plaintiff's legal team has voluntarily reduced its requested attorney fees from $300 million to $187.5 million, related to the prior $1.5 billion settlement with AI giant Anthropic.

This adjustment is not an act of generosity by the law firm, but a response to pressure from both the court and the defendant. Previously, Anthropic and the presiding judge had opposed the $300 million "sky-high" fee, arguing the amount was excessive and involved an unreasonable cost-sharing structure.

The $1.5 Billion Settlement: The Price of AI Training

As the developer of the Claude model series, Anthropic agreed last August to pay $1.5 billion to settle claims it used hundreds of thousands of pirated books to train its AI. This stands as the largest copyright class-action settlement to date and sets an industry precedent: Anthropic committed to destroying the relevant pirated datasets and ensuring future commercial models avoid such materials.

Under the agreement, each copyright holder of a protected work will receive over $3,000 in compensation. This high payout is viewed as a significant victory for traditional content creators against the AI industry's practice of "free feeding" on copyrighted material.

The Attorney Fee Dispute: Who Truly Serves the "Collective Interest"?

The fee reduction centers on the legal team's internal structure. The original $300 million request included $75 million for three other law firms not officially designated as class counsel. The judge clarified that the lead firm cannot unilaterally "appoint others" to share responsibility and fees. In new filings, the attorneys acknowledged accepting the court's guidance, stating the revised $187.5 million request—approximately 12.5% of the settlement fund—will be based solely on the work of the officially appointed lead counsel.

April Hearing: The Settlement's Final Stage

The U.S. District Court for the Northern District of California has scheduled an April 23 hearing to decide on final approval of the revised settlement. With this key hurdle addressed, this protracted AI copyright case is poised to conclude.

For the broader AI industry, this substantial settlement is not merely a payment for past actions but also a ticket into a "compliance era." As legal boundaries clarify, the future model for collaboration between AI developers and copyright holders is shifting from "use first, litigate later" to "secure permission first, then train."