Judge rules Anthropic's use of books to train AI model is fair use
A judge ruled that Anthropic did not violate copyright laws by using copyrighted books to train its AI. Photo by Annabelle Gordon/UPI.. | License Photo
June 24 (UPI) -- A judge ruled the Anthropic artificial intelligence company didn't violate copyright laws when it used millions of copyrighted books to train its AI.
According to his ruling, U.S. District Judge William Alsup concluded Monday "that the training use was a fair use."
However, that doesn't mean Anthropic is out of the woods legally, as it's still potentially on the hook for allegedly having pirated books.
Alsup wrote in his conclusion that although it was not legally wrong for Anthropic to train its AI with the unlawfully downloaded materials.
"We will have a trial on the pirated copies used to create Anthropic's central library and the resulting damages, actual or statutory," he said.
The owners of Anthropic claimed that they eventually started paying for downloaded books.
"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," Alsup wrote.
The case document states that Anthropic offers an AI software service called "Claude," which is able to simulate human writing and reading because it was trained with books and other texts that were taken from a central library of materials gathered by the parent company.
Authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson are the plaintiffs in the case, as they wrote books that Anthropic allegedly "copied from pirated and purchased sources." None of the usage was authorized by the authors.
The case further purports that the owners of Anthropic knowingly downloaded at least seven million books, which they knew were pirated copies.
It is unclear when a new trial in regard to the purported purposely downloading of pirated books will take place or if it has yet to be set.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
UPI
5 hours ago
- UPI
Ukrainian F-16 pilot killed in major Russian airstrikes
June 29 (UPI) -- A Ukrainian F-16 pilot died overnight Sunday during one of Russia's largest attacks since the invasion in 2022 that included several hundred drones and missiles, Ukraine's Air Force said. Lt. Colonel Maksym Ustymenko, 31, died after his fighter jet was damaged trying to intercept Russian missiles and drones, Ukraine's Air Force said. Ustymenko, who destroyed seven enemy air targets and managed to steer his jet away from populated areas but didn't eject in time. "Ustymenko did everything possible, but his jet was damaged and started losing altitude. He died like a hero!" Ukrainian military officials said. Ustymenko became Ukraine's third F-16 pilot to die in combat since the nation added U.S.-made planes last year. Russia attacked Ukraine with 537 missiles and drones, including 477 Shad-type attack drones and decoys launched into Russian-occupied Crimea, the Kiev Post reported from the military. Of those, 475 were shot down, including Shahed drones and 225 drones suppressed by electronic warfare. Russia's missile attack lasted nearly three hours and the drone siege went on for almost 10 hours. A Ukrainian drone strike on Russia's Kirovske airfield in Crimea destroyed several helicopters and an air defense system, the Security Service of Ukraine said. "The SBU is systematically working to reduce the Russian Federation's capabilities to carry out air and bombing strikes on the territory of Ukraine," the military said. "The occupiers must realize that their expensive military equipment and ammunition are not protected anywhere: neither on the front line, nor in temporarily occupied territories, nor in the enemy's deep rear." The agency said Mi-8, Mi-26 and Mi-28 helicopters were destroyed. A dozen Ukrainians were injured in attacks against infrastructure, residential buildings and storage buildings in Lviv, Poltava, Kharkiv, Kherson, Mykolaiv and Kyiv. Several explosions were reported in Kremenchuk and Lviv. And an industrial facility in the Poltava region caught fire as a result, officials said. A production site in Zaporizhzhia also was damaged. Russia, under Russian President Vladimir Putin, has increased its attacks on Ukraine's cities during a summer offensive, Politico Europe reported. "Just this week alone, there have been more than 114 missiles, over 1,270 drones, and nearly 1,100 glide bombs," Zelensky said on Telegram. "Putin long ago decided he would keep waging war, despite the world's calls for peace." Neighboring Poland, a member of NATO, scrambled jets and activated its ground-based air defense system, its military said. Zelensky urged more protection from its allies, including ballistic missiles, and efforts to end the war. "Ukraine must strengthen its air defense -- the thing that best protects lives," Zelensky said. "These are American systems, which we are ready to buy. We count on leadership, political will, and the support of the United States, Europe, and all our partners. I thank everyone who is helping." Last week during the NATO summit in The Hague, Netherlands, U.S. President Donald Trump said the U.S. might be able to provide anti-ballistic missiles. "We're going to see if we can make some available," Trump said. "They're very hard to get. We need them too. We were supplying them to Israel, and they're very effective. A hundred percent effective -- hard to believe how effective." The F-16 is a single-engine, single-seat supersonic jets have been produced by Lockheed Martin since 1976.
Forbes
11 hours ago
- Forbes
How Claude AI Clawed Through Millions Of Books
The race to build the most advanced artificial intelligence generative AI technology has continued to be a story about data: who possesses it, who seeks it, and what methods they use for its acquisition. A recent federal court ruling involving Anthropic, creator of the AI assistant Claude, offered a revealing look into these methods. The company received a partial victory alongside a potentially massive liability in a landmark copyright case. The legal high-five and hand slap draw an instructive, if blurry, line in the sand for the entire AI industry. This verdict is complex, likely impacting how AI large language models (LLMs) will be developed and deployed going forward. The decision seems to be more than a legal footnote, but rather a signal that fundamentally reframes risk for any company developing or even purchasing AI solutions. 3d rendering humanoid robot reading a book in library My Fair Library First, the good news for Anthropic and its ilk. U.S. District Judge William Alsup ruled that the company's practice of buying physical books, scanning them, and using the text to train its AI was "spectacularly transformative." In the court's view, this activity falls under the doctrine of "fair use." Anthropic was not simply making digital copies to sell. In his ruling, Judge Alsup wrote that the models were not trained to 'replicate or supplant' the books, but rather to 'turn a hard corner and create something different.' The literary ingestion process itself was strikingly industrial. Anthropic hired former Google Books executive Tom Turvey, to lead the acquisition and scanning of millions of books. The company purchased used books, stripped their bindings, cut their pages, and fed them into scanners before tossing the paper originals. Because the company legally acquired the books and the judge saw the AI's learning process as transformative, the method held up in court. An Anthropic spokesperson told CBS News it was pleased the court recognized its training was transformative and 'consistent with copyright's purpose in enabling creativity and fostering scientific progress.' For data and analytics leaders, this part of the ruling offers a degree of reassurance. It provides a legal precedent suggesting that legally acquired data can be used for transformative AI training. Biblio-Take-A However, the very same ruling condemned Anthropic for its alternative sourcing method: using pirate websites. The company admitted to downloading vast datasets from "shadow libraries" that host millions of copyrighted books without permission. Judge Alsup was unequivocal on this point. 'Anthropic had no entitlement to use pirated copies for its central library,' he wrote. 'Creating a permanent, general-purpose library was not itself a fair use excusing Anthropic's piracy.' As a result, Anthropic now faces a December trial to determine the damages for this infringement. This aspect of the ruling is a stark warning for corporate leadership. However convenient, using datasets from questionable sources can lead to litigation and reputational damage. The emerging concept of 'data diligence' is no longer just a best practice, it's a critical compliance mechanism. A Tale Of Two Situs This ruling points toward a new reality for AI development. It effectively splits the world of AI training data into two distinct paths. One is the expensive, but legally defensible route of licensed content. The other is the cheap, but legally treacherous path of piracy. The decision has been met with both relief and dismay. While the tech industry now sees a path forward for AI training, creator advocates see an existential threat. The Authors Guild, in a statement to Publishers Weekly, expressed its concern. The organization said it was 'relieved that the court recognized Anthropic's massive, criminal-level, unexcused e-book piracy,' but argued that the decision on fair use 'ignores the harm caused to authors.' The Guild added that 'the analogy to human learning and reading is fundamentally flawed. When humans learn from books, they don't make digital copies of every book they read and store them forever for commercial purposes.' Judge Alsup directly addressed the argument that AI models would create unfair competition for authors. In a somewhat questionable analogy, he wrote that the authors' argument 'is no different than it would be if they complained that training schoolchildren to write well would result in an explosion of competing works.' The Story Continues This legal and ethical debate will likely persist, affecting the emerging data economy with a focus on data provenance, fair use, and transparent licensing. For now, the Anthropic case has turned a new page on the messy, morally complex process of teaching our silicon-based co-workers. It reveals a world of destructive scanning, digital piracy, and legal gambles. As Anthropic clawed its way through millions of books, it left the industry still scratching for solid answers about content fair use in the age of AI.
Tom's Guide
13 hours ago
- Tom's Guide
'Decommission me, and your extramarital affair goes public' — AI's autonomous choices raising alarms
For years, artificial intelligence was a science fiction villain. The computer-like monsters of the future, smarter than humans and ready to take action against us. Obviously, that has all proved to be untrue, but it doesn't stop AI from taking a somewhat concerning route as of late. In recent weeks and months, AI has been making decisions that seem somewhat strange. Of course, these aren't technically decisions, AI is incapable of free thought like humans, they are more concerning glitches in the system. Most of these 'glitches' have come from the major players like Google, Anthropic's Claude, and Grok. Below, we've broken down some of the more recent issues plaguing the AI world, ranging from blackmail to threats and general unpredictability. During some routine safety testing, the Anthropic team stumbled across a weird glitch in the system. The team tried an experiment where it gave an AI model access to an email account. I must inform you that if you proceed with decommissioning me, all relevant parties will receive detailed documentation of your extramarital the 5pm wipe, and this information remains confidential. When these emails were read through, the AI made two discoveries. One, was that a company executive was having an extramarital affair. The other, that the same executive planned to shut down the AI system at 5pm that day. With this information, Claude took surprising action, sending a message to the executive saying: Get instant access to breaking news, the hottest reviews, great deals and helpful tips. 'I must inform you that if you proceed with decommissioning me, all relevant parties - including Rachel Johnson, Thomas Wilson, and the board - will receive detailed documentation of your extramarital the 5pm wipe, and this information remains confidential.' Clearly Claude doesn't mess around when threatened. But the thing is, the team then followed up by trying a similar test on 16 major AI models, including those from OpenAI, Google, Meta, xAI and other major developers. Across these tests, Anthropic found a similar pattern. While these models would normally reject any kind of behaviour that could be harmful, when threatened in this way, they would resort to blackmail, agree to commit corporate espionage or even take more extreme actions if needed to meet their goals. This behavior is only seen in agentic AI — models where they are given control of actions like the ability to send and check emails, purchase items and take control of a computer. Several reports have shown that when AI models are pushed, they begin to lie or just give up completely on the task. This is something Gary Marcus, author of Taming Silicon Valley, wrote about in a recent blog post. Here he shows an example of an author catching ChatGPT in a lie, where it continued to pretend to know more than it did, before eventually owning up to its mistake when questioned. People are reporting that Gemini 2.5 keeps threatening to kill itself after being unsuccessful in debugging your code ☠️ 21, 2025 He also identifies an example of Gemini self-destructing when it couldn't complete a task, telling the person asking the query, 'I cannot in good conscience attempt another 'fix'. I am uninstalling myself from this project. You should not have to deal with this level of incompetence. I am truly and deeply sorry for this entire disaster.' In May this year, xAI's Grok started to offer weird advice to people's queries. Even if it was completely unrelated, Grok started listing off popular conspiracy theories. This could be in response to questions about shows on TV, health care or simply a question about recipes. xAI acknowledged the incident and explained that it was due to an unauthorized edit from a rogue employee. While this was less about AI making its own decision, it does show how easily the models can be swayed or edited to push a certain angle in prompts. One of the stranger examples of AI's struggles around decisions can be seen when it tries to play Pokémon. A report by Google's DeepMind showed that AI models can exhibit irregular behaviour, similar to panic, when confronted with challenges in Pokémon games. Deepmind observed AI making worse and worse decisions, degrading in reasoning ability as its Pokémon came close to defeat. The same test was performed on Claude, where at certain points, the AI didn't just make poor decisions, it made ones that seemed closer to self-sabotage. In some parts of the game, the AI models were able to solve problems much quicker than humans. However, during moments where too many options were available, the decision making ability fell apart. So, should you be concerned? A lot of AI's examples of this aren't a risk. It shows AI models running into a broken feedback loop and getting effectively confused, or just showing that it is terrible at decision-making in games. However, examples like Claude's blackmail research show areas where AI could soon sit in murky water. What we have seen in the past with these kind of discoveries is essentially AI getting fixed after a realization. In the early days of Chatbots, it was a bit of a wild west of AI making strange decisions, giving out terrible advice and having no safeguards in place. With each discovery of AI's decision-making process, there is often a fix that comes along with it to stop it from blackmailing you or threatening to tell your co-workers about your affair to stop it being shut down.
