Anthropic Settles Copyright Lawsuit for $1.5 Billion, Key IPO Legal Risks Largely Eliminated

TradingKey - On July 21, Eastern Time, the U.S. District Court for the Northern District of California formally approved a final $1.5 billion class-action copyright settlement agreement between artificial intelligence giant Anthropic and a group of authors, officially ending the dispute over its use of pirated books to train its Claude LLM. The closing of this deal completely shattered the record for the largest copyright settlement in U.S. history.

The lawsuit originated in 2024 when authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson filed the Bartz v. Anthropic case, alleging that Anthropic downloaded over 7 million works from pirated book repositories such as Library Genesis and Pirate Library Mirror to train its Claude series of large language models. In September 2025, Judge William Alsup ruled that training AI on books does not violate U.S. copyright law, but determined that Anthropic must stand trial for its use of pirated content.

Anthropic subsequently agreed to a $1.5 billion settlement to avoid further protracted litigation. Under the terms of the settlement, copyright owners of each eligible book will receive approximately $3,000 in compensation, and Anthropic must destroy all copies of files obtained from the pirated book repositories.

Notably, Anthropic did not admit to any wrongdoing in the settlement. The settlement only resolves claims related to past downloading and training activities, while rights holders retain the right to file separate lawsuits for future model training or generative outputs. Some publishers and authors have already opted out of the settlement to file independent lawsuits of their own.

In June of this year, Anthropic confidentially filed for an IPO, with its valuation anchored at $965 billion, and the market is closely watching its listing process. The $1.5 billion settlement amount represents approximately 3.4% of Anthropic's annualized revenue (roughly $44 billion), but spending this money resolves the litigation risks stemming from past downloading and training activities.

In addition to the copyright litigation, Anthropic has recently restored full access to Claude Fable 5 and Mythos 5, following the resolution of compliance issues previously related to export control directives from the U.S. Department of Commerce.

The end of the copyright lawsuit does not mean regulatory risks are over. The U.S. Executive Order on AI Safety requires frontier AI models to submit evaluations to the government before public beta testing, and the AI Standards and Innovation Center under the Department of Commerce is building a new framework covering model testing, data sources, and export controls. During congressional hearings, some lawmakers directly demanded that Anthropic disclose its training data sources and model evaluation reports. Once passed, such legislative requirements would impose substantive constraints on all AI companies.

As of press time, Anthropic has not provided further comment on the settlement's impact on its IPO timeline. What is certain is that after the removal of the biggest legal variable, the market's scrutiny of this company, valued at $965 billion, will return to the most core question: can its technical roadmap and revenue growth sustain this valuation? This is also the critical proposition that the capital markets will need to answer next.