AI Models, Copyright, & Books: Navigating the Legal Minefield

You probably know by now that the AI models powering ChatGPT, Gemini, Claude, and other chatbots are trained on seemingly infinite databases of published works, containing hundreds of millions of books, online articles, academic papers, and basically anything you can find on the internet. Most published authors have, without their knowledge or consent, contributed to the development of the same AI tools that threaten to undermine their livelihoods. That seems illegal, right?

Key Takeaways

• AI Training vs. Piracy:A landmark $1.5 billion settlement against Anthropic clarified that while AI training on copyrighted material *may* be lawful, obtaining that material through piracy is not. This distinction offers a significant advantage to AI companies.
• Fair Use in Flux:The core of AI copyright disputes hinges on the “fair use” doctrine, particularly whether AI’s ingestion of data is “transformative.” Courts are currently divided, often ruling against AI if it directly competes with the original copyrighted work.
• The Evolving Legal Landscape:With copyright law largely unchanged since 1976, judges are grappling with unprecedented questions. Early rulings are influential but not definitive, signaling a complex and protracted legal battle for the future of AI and creative industries.

The reality isn’t that simple, and the legal battlegrounds forming around artificial intelligence and intellectual property are anything but clear-cut. As AI rapidly integrates into our daily lives and creative processes, the foundational principles of copyright are being tested in ways unimaginable just a few years ago.

“I think one of the issues with this entire area of law and this entire area of technology is there’s a lot going on,” Cathy Gellis, an attorney with expertise in intellectual property, copyright, and technology, told TechCrunch. “It’s very complex and there are a lot of raw feelings about what is happening, both for and against.” This complexity stems from the rapid advancement of AI against the backdrop of laws conceived in a pre-digital, let alone pre-AI, era.

The Anthropic Ruling: A Double-Edged Sword for Authors

Last year, in one of the first rulings of its kind, Judge William Alsup ordered Anthropic to pay a mammoth $1.5 billion copyright settlement to a group of writers whose works were used to train the company’s AI models. At face value, this seemed like a moral victory favoring authors, a strong signal that AI companies would be held accountable for leveraging creators’ work. However, the true nuance of Judge Alsup’s decision reveals a more complex reality.

Crucially, Judge Alsup actually ruled that Anthropic’s *AI training itself* was lawful. What Alsup penalized Anthropic for was *pirating* these books from illegal online shadow libraries. The distinction is vital: the court found the *method of acquisition* illegal, not the *act of training* on the acquired data.

“Like any reader aspiring to be a writer, Anthropic’s LLMs trained upon works not to race ahead and replicate or supplant them — but to turn a hard corner and create something different,” the judge wrote, drawing a parallel between an LLM ingesting trillions of words and a human writer’s study of literature. This analogy suggests a judicial inclination to view AI training as a form of “reading” or “learning,” rather than direct “copying” in the traditional copyright sense.

Gellis thinks the ruling is, in many ways, more advantageous for AI companies than it might initially appear. From a business perspective, what’s a $1.5 billion fine to a company projecting about $200 billion in annual revenue by 2028? It’s a cost of doing business, rather than a prohibitive roadblock to their core technology.

“I think it is generally good news for AI training that he looked at what was going on and really sort of thought it analogous to reading a copyrighted work as opposed to copying a copyrighted work,” Gellis said. “Copyright law hinges on copying, but it doesn’t hinge on using the work or experiencing the work, consuming the work, reading the work.” This interpretation opens a wide door for AI developers, potentially legitimizing the vast ingestion of data needed for sophisticated models, as long as the data is legally sourced.

Outdated Laws, Modern Problems: The Fair Use Conundrum

The fundamental challenge in these cases is that copyright law hasn’t been significantly updated since 1976. This means that judges are tasked with interpreting guidelines from nearly 50 years ago to confront legal questions that have the potential to shape the entire future of the AI industry. The sheer scale and speed of AI development expose the limitations of existing legal frameworks.

“Everybody is very worried right now because the law is all over the place, and it’s because of this question,” Jason Henderson, Senior Attorney and Founder of the IP & Media Practice at JWL International, told TechCrunch. “They know that the AI model has been trained on so much stuff, and the law has not really caught up to that question.”

These complex questions often hinge on the doctrine of fair use – a crucial carve-out of copyright law that allows for the limited use of copyrighted materials without explicit permission from the rights holder. Fair use exists to protect the ability to comment, criticize, parody, educate, and innovate upon existing works. Judges consider specific factors when deciding if something qualifies as fair use, including:

• The purpose and character of the use (e.g., commercial vs. non-profit, transformative vs. derivative).
• The nature of the copyrighted work (e.g., factual vs. creative).
• The amount and substantiality of the portion used in relation to the copyrighted work as a whole.
• The effect of the use upon the potential market for or value of the copyrighted work.

“Copyright is always about protecting and growing the market,” Henderson noted. “The courts are kind of all over the place in their reasoning [in AI cases]. What’s tending to win is if what you’re doing is you’re training on somebody’s property because your purpose is to directly compete, then the courts will frown on it… If what you’re doing is not going to compete, then the courts are tending to find ways that it will be okay.”

Competition as a Litmus Test: Thomson Reuters vs. Ross Intelligence

Henderson is referencing a significant case in which the media and technology company Thomson Reuters sued the research firm Ross Intelligence. Ross was accused of copying Thomson Reuters’ content in order to build a directly competing, AI-based legal research platform. This case provided a clearer example of when courts might draw the line.

“Ross’s use is not transformative because it does not have a ‘further purpose or different character’ than Thomson Reuters’s,” Judge Stephanos Bibas wrote last year. In this instance, Judge Bibas determined that training on Reuters’ content to create a platform that would directly compete with it was not fair use. This decision highlights that if the AI’s output directly supplants the original market for the copyrighted work, courts are more likely to rule against the AI company.

While authors could potentially argue that generative AI chatbots are competing with them by using their works to generate new, synthetic books or articles, that specific argument—that AI *output* constitutes direct market competition with human authors—has not yet definitively prevailed in court. The distinction between training an AI to understand and generate new content versus directly replicating or competing with existing content remains a key battleground.

Beyond Training: The AI Authorship Dilemma

When it comes to the relationship between AI and copyright, Gellis finds it helpful to narrow down what we’re actually talking about – the way we think about copyright in terms of AI training is quite different from how we think about copyrighting AI-generated content. The latter introduces another layer of complexity to the legal landscape.

In one notable case, Thaler v. Perlmutter, the court ruled that if a work is 100% AI-generated, it’s not copyrightable under current U.S. law, as copyright requires human authorship. This opens a whole new can of worms: how can we definitively prove whether or not a work was generated using AI, and if so, how do we know what percentage of it was created or assisted with AI? The lines between human creativity and algorithmic contribution are blurring rapidly.

“If you write your novel in [Microsoft] Word and run spell check, we kind of feel comfortable with the idea of saying that Word does not own your novel,” Gellis said. But AI’s role can be far more significant than a spell checker, raising fundamental questions about creativity, ownership, and the very definition of an “author.” “[AI] is forcing us to look at a whole bunch of decisions that we kind of ignored for a while.”

The Unfolding Legal Saga

Most major AI companies are still lodged in pending litigation over these multifaceted issues, which means that we won’t have a definitive, universally applicable solution to these problems any time soon. The legal landscape is highly dynamic, with each new ruling potentially influencing subsequent cases and interpretations.

“What you are seeing is that the initial opening volleys are being influential, and that influence itself could be undone if other courts decide different things, and it’ll take later states of litigation to figure out which one will prevail,” Gellis said. “But in the meantime, all these decisions are shaping everything that’s happening. It would be kind of foolish for the AI companies to ignore them.” The ongoing legal skirmishes are not just about monetary damages; they are about setting precedents that will define the future boundaries of AI innovation and intellectual property rights.

The Bottom Line

The legal battle over AI and copyright is a complex, evolving saga where outdated laws clash with cutting-edge technology. While early rulings suggest AI training itself might be defensible under a “transformative use” lens, the acquisition of data and the potential for direct market competition remain critical points of contention. As courts navigate these unprecedented waters, the outcomes will undoubtedly reshape not only the AI industry but also the foundational principles of creativity, authorship, and intellectual property in the digital age. Creators, developers, and legal experts alike are watching closely as the future of both innovation and artistic rights hangs in the balance.