Artists are taking AI companies to court over unauthorized use of their work in training data, and some are winning meaningful victories that could reshape how the industry operates. From illustrators to novelists to musicians, creative professionals are filing lawsuits against major tech firms including Google, Meta, Anthropic, and Stability AI, arguing that their copyrighted works were pirated and fed into AI models without permission or compensation.

Why Are Artists Suing AI Companies?

The catalyst for many lawsuits came when The Atlantic published a searchable dataset revealing which works had been used to train AI systems. Author Kirk Wallace Johnson discovered that his nonfiction books, "The Feather Thief" and "The Fishermen and the Dragon," which he spent five to six years researching and writing, had been included without his knowledge. He described feeling a "cocktail" of emotions: "anger over the brazenness of the theft, worry over what this means for writers, and a healthy thirst for revenge on these massive corporations that have become galactically wealthy" using his intellectual property.

The concern extends beyond famous authors. Novelist Richard Kadrey and others who sued Meta argued that the real threat isn't to household names like Agatha Christie, but to working artists. "AI could write a mediocre book," Johnson explained, "and there are tons of authors and screenwriters that live in that space. And it's no judgment to them. They're servicing a marketplace." The fear is that AI-generated content could flood markets where mid-list creators depend on sales to sustain their careers.

What Legal Arguments Are Artists Using?

Artists have pursued multiple legal strategies, though copyright infringement remains the primary claim. Some cases focus on different angles entirely. Musician Sam Kogon, the lead plaintiff in a suit against Google's Lyria AI music engine, is accusing the company of violating its own terms of service rather than relying solely on copyright law. Kogon's lawyers argue that Google improperly used its Content ID system and YouTube data to train Lyria and ProducerAI without proper authorization.

Kogon views the situation as "pure bait and switch." YouTube's terms of service are dense, compulsory agreements that users cannot negotiate. "Technology that wasn't even invented, and wasn't even a glimmer in anyone's eye at the time of putting your things on YouTube, is now fair game," he argued. Entertainment and IP lawyer Krystle Delgado agreed, noting that "I don't think that anyone uses YouTube thinking that you are giving the rights to remake your content".

The legal landscape remains uncertain because courts are still defining what constitutes fair use in the AI context. Fair use typically permits material to be used without permission for research, teaching, or commentary. AI companies have argued that learning from copyrighted works falls within this doctrine, but artists and their lawyers dispute this interpretation.

How Are Courts Responding to These Cases?

Results have been mixed. Illustrator Sarah Andersen was among the first to directly challenge AI giants, filing a class action suit against Stability, Midjourney, DeviantArt, and Runway AI in January 2023, just months after Stable Diffusion and Midjourney launched. The case has been crawling through the court system since then. In the Meta lawsuit, a judge dismissed many of the authors' initial claims for failing to show evidence of market harm, though a narrower set of claims survived.

Despite setbacks, artists remain cautiously optimistic. Many believe their efforts will help guide courts toward establishing legal guardrails for AI development. Andersen described her webcomic "Sarah's Scribbles" as a "complex culmination of my education, the comics I devoured as a child, and the many small choices that make up the sum of my life," and felt "violated" when she discovered it had been used to train AI models without her consent.

What's Driving the Broader IP Litigation Explosion?

The surge in AI-related copyright disputes reflects a dramatic shift in how intellectual property is valued and contested. According to recent data, AI-related copyright cases in the United States skyrocketed from just 16 cases in 2023 to 70 cases in 2025, representing a more than fourfold increase in just two years. This explosion is part of a larger transformation: intangible assets like proprietary data, technical inventions, brands, and creative works now represent over 90 percent of market value in the S&P 500, compared to just 17 percent in 1975.

The broader IP litigation landscape has also diversified significantly. Two decades ago, patent disputes dominated intellectual property courts. Today, cases involving copyrights, trademarks, and trade secrets have grown substantially. U.S. IP case filings have surged from fewer than 3,000 annual cases in the mid-2000s to nearly 19,000 by 2025.

What Are the Key Concerns About AI Distillation?

Beyond traditional copyright disputes, a new threat has emerged: AI distillation. This practice involves training a new AI model by running thousands of prompts through an existing, more advanced model to see what good answers look like. It's essentially a shortcut that leverages the work of others without building from scratch.

The Trump administration has accused Chinese startup Moonshot AI of using distillation to develop its Kimi K3 model from Anthropic's work. Anthropic has warned that distillation by Chinese companies poses a national security risk, particularly for models that could be misused in cyberattacks or biological research. Treasury Secretary Scott Bessent stated that the administration had evidence many Chinese AI models were engaging in distillation and could impose sanctions.

The legal status of distillation remains untested in court. While companies could potentially argue that distillation constitutes copyright infringement, trade secret theft, or terms of service violations, none of these theories has been definitively established. The irony is not lost on observers: American AI companies that built their models by training on copyrighted works without permission now face accusations that others are doing the same to them.

Steps Artists Can Take to Protect Their Work

• Document Your Work: Keep detailed records of creation dates, drafts, and publication history to establish ownership and originality when pursuing legal claims against AI companies.
• Monitor Training Datasets: Regularly search publicly available datasets like those published by The Atlantic to determine whether your work has been included in AI training without authorization.
• Join Collective Actions: Consider participating in class action lawsuits with other artists, as these cases have greater resources and visibility than individual suits, increasing chances of meaningful settlements.
• Review Platform Terms: Carefully examine the terms of service for any platform where you upload your work, as companies may claim broad rights to use content for AI training purposes.
• Consult IP Lawyers: Seek guidance from entertainment and intellectual property attorneys who specialize in AI disputes to understand your legal options and the strength of potential claims.

Illustrator Sarah Andersen captured the frustration many artists feel about the approach taken by AI companies. "This does not seem to be a bus driven by a bunch of sane sober thinking people, and we're all stuck in it," she said, describing her work as being reduced "to an algorithm".

Author Andrea Bartz, the lead plaintiff in Susman Godfrey's suit against Anthropic, expressed similar emotions. "I felt violated, shocked, alarmed," she told The Verge. "I had a big emotional response to seeing that something I'd worked on for so many years and poured my heart and soul into was just one of hundreds of thousands or maybe millions of books that these Big Tech companies had just stolen for training their algorithm".

The outcome of these cases will likely determine whether AI companies must negotiate licenses with creators, pay licensing fees, or implement opt-out mechanisms before training on copyrighted works. For now, the legal landscape remains in flux, with courts grappling with age-old questions about market substitution and how traditional licensing structures must adapt to models trained on existing human works.