A federal judge in San Francisco has delivered a split decision in a closely-watched copyright lawsuit against artificial intelligence company Anthropic, ruling that using copyrighted books to train AI systems constitutes fair use while simultaneously ordering the company to face trial for downloading millions of pirated works.
The groundbreaking decision by U.S. District Judge William Alsup represents the first time a federal court has addressed fair use in the context of generative AI training. According to Reuters, Alsup ruled Monday that Anthropic’s use of books by authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson to train its Claude language model was “exceedingly transformative” and therefore protected under fair use doctrine. However, the judge also determined that as reported by TechCrunch, Anthropic’s creation and storage of a “central library” containing over 7 million pirated books violated copyright law.
The Legal Landscape for AI Companies
The ruling comes amid a surge of copyright litigation against AI companies operating in the San Francisco Bay Area. Thirty-right copyright lawsuits have been filed against AI companies in the United States, with the Northern District of California handling the most cases at 20 total. The concentration of these cases in San Francisco reflects the city’s status as a hub for both AI development and creative industries.
Judge Alsup, a veteran jurist known for his tech-savvy approach to complex cases, drew on an analogy that The Fashion Law noted when he wrote that Anthropic’s AI models “trained upon works not to race ahead and replicate or supplant them — but to turn a hard corner and create something different,” comparing the process to “any reader aspiring to be a writer.”
Pattern of Mixed Results for Tech Companies
The Anthropic decision follows a pattern of nuanced rulings in AI copyright cases before Northern District of California judges. Earlier this year, Reuters reported that Judge Vince Chhabria expressed skepticism about Meta’s fair use arguments in a similar case, warning that AI could “obliterate” markets for original works. The varying judicial perspectives highlight the evolving legal landscape as courts grapple with unprecedented questions about AI training and copyright law.
Legal experts say the Anthropic ruling could significantly influence dozens of pending cases against other major AI companies including OpenAI, Microsoft, and Meta Platforms. The Copyright Alliance observed that while no final decisions have been reached in the broader litigation landscape, 2024 saw several developments that hint at how courts are leaning on fair use questions.
The Plaintiffs and Their Claims
The case was brought by three established authors who alleged that Anthropic built its business model on what they characterized as “large-scale theft.” Andrea Bartz, author of novels including “The Lost Night” and “The Herd,” joined forces with Charles Graeber, known for “The Good Nurse: A True Story of Medicine, Madness, and Murder,” and Kirk Wallace Johnson, author of “The Fisherman and the Dragon: Fear, Greed, and a Fight for Justice on the Gulf Coast.” The Hollywood Reporter detailed how the proposed class action sought to represent other authors whose books were used as training data.
In their lawsuit filed in August 2024, the authors argued that according to their legal team at Lieff Cabraser, Anthropic “intentionally downloaded known pirated copies of books from the internet, made unlicensed copies of them, and then used those unlicensed copies to digest and analyze the copyrighted expression—all for its own commercial gain.”
Trial Set for December on Piracy Claims
While Alsup dismissed the core copyright infringement claims related to AI training, he ordered a trial in December to determine damages for Anthropic’s use of pirated books. The judge noted that Associated Press reported, “That Anthropic later bought a copy of a book it earlier stole off the internet will not absolve it of liability for the theft but it may affect the extent of statutory damages.”
Under U.S. copyright law, willful infringement can result in statutory damages of up to $150,000 per work. With Anthropic allegedly maintaining a library of over 7 million pirated books, the potential financial exposure could be substantial.
Broader Implications for the AI Industry
The ruling represents a significant moment for the artificial intelligence industry, which has increasingly relied on vast datasets of copyrighted material to train sophisticated language models. San Francisco Public Press noted that California creatives have been rallying behind state-level AI regulations to protect their artwork from unauthorized use in training datasets.
San Francisco-based Anthropic, headquartered in the financial district, has positioned itself as a leader in AI safety and ethics. The company was founded in 2021 by former OpenAI executives Dario and Daniela Amodei and has received significant investments from Amazon and Google. The Real Deal reported that Anthropic has been expanding its San Francisco presence, including a pending sublease of Slack’s 230,000-square-foot headquarters at 500 Howard Street.
In a statement following the ruling, Anthropic expressed satisfaction that the court recognized AI training as transformative and consistent with copyright law’s purpose of enabling creativity and fostering scientific progress. The company did not address the piracy allegations in its public response.
Rising Legal Costs and Defensive Strategies
The proliferation of copyright challenges has prompted AI companies to invest heavily in legal defense. Industry tracking shows that OpenAI alone has hired seven trial attorneys from the prestigious firm Keker, Van Nest & Peters to represent it across multiple lawsuits.
Meanwhile, Reuters previously reported that Anthropic faced additional scrutiny in a separate case when a judge ordered the company to respond to allegations that an expert witness cited a fabricated academic article, potentially created by AI, in a copyright defense filing.
Industry Response and Future Outlook
The mixed nature of Alsup’s ruling reflects the complex legal questions surrounding AI development and intellectual property rights. While tech companies celebrated the fair use finding, creators and copyright holders emphasized that the piracy ruling validates their concerns about unauthorized use of their works.
The decision arrives as Congress and state legislatures consider new frameworks for regulating AI development and protecting creator rights. California lawmakers are considering legislation that would require AI companies to disclose what copyrighted materials they use in training datasets, a measure supported by creative industry groups.
As the AI industry continues to evolve rapidly, the Anthropic ruling provides initial guidance while leaving many questions unresolved. The December trial on piracy damages will be closely watched as another indicator of how courts balance innovation incentives with intellectual property protections in the digital age.
Copyright for syndicated content belongs to the linked Source link
