Artists Win Legal Ground Against AI Training Without Consent
A wave of copyright lawsuits from creators against major AI companies is forcing courts to define fair use in the age of generative models.
Last verified:
A growing cohort of creators — authors, illustrators, and musicians — is taking AI companies to court over unauthorized use of their work in training datasets. According to The Verge, some have already secured victories, while courts begin to establish legal boundaries around fair use in the generative AI era.
High-Profile Authors and Artists Launch Coordinated Legal Challenge
Author Kirk Wallace Johnson discovered his published books — including The Feather Thief and The Fishermen and the Dragon, works he spent five to six years researching and writing — included in The Atlantic’s searchable training-data dataset without his consent. According to The Verge, Johnson joined a lawsuit led by Susman Godfrey against Anthropic on behalf of multiple authors. His decision reflects a broader pattern: dozens of creators across disciplines are now filing suits against major AI vendors on copyright grounds.
Illustrator Sarah Andersen, whose webcomic Sarah’s Scribbles appears in training datasets, filed a class-action suit in January 2023 against Stability AI, Midjourney, DeviantArt, and Runway AI alongside visual artists Karla Ortiz and Kelly McKernan. The Verge reports that the case has been working through the court system for years. Other lawsuits have followed, targeting Google, Meta, Anthropic, and the AI music generator Suno.
Mixed Outcomes and Emerging Fair-Use Precedent
According to The Verge, artists have experienced both wins and setbacks in these disputes. Some cases have settled quickly; others have dragged on for years. The outcomes often hinge on how courts interpret fair use — a legal doctrine that has not yet stabilized in the context of large-scale model training.
The definition of fair use remains contested. Courts are now beginning to establish precedent, though creatives acknowledge uncertainty about which arguments will ultimately prevail. Kirk Wallace Johnson expressed skepticism about the broader AI industry’s judgment, telling The Verge the situation feels chaotic: “This does not seem to be a bus driven by a bunch of sane sober thinking people, and we’re all stuck in it.”
Why This Matters
These lawsuits are shaping regulatory and legal frameworks that will govern AI training for years. If courts rule that fair use does not permit scraping copyrighted works without compensation or consent, companies will need to secure licensing agreements or rebuild models with licensed or public-domain data. This could raise the cost of training and shift competitive advantage toward companies with deep pockets for licensing. For individual creators, the outcomes will determine whether AI companies must negotiate with rights holders before publication or face retroactive damages.
The stakes extend beyond copyright: the cases are also testing whether terms-of-service violations can serve as independent claims against AI vendors — a question that could ripple into consumer protection and data privacy law.
Frequently Asked Questions
What are artists suing AI companies for?
Primarily copyright infringement — companies trained models on copyrighted works without permission or compensation. Some suits also challenge violations of terms of service.
Which AI companies are being sued?
According to The Verge, the defendants include Anthropic, Meta, Google, Stability AI, Midjourney, DeviantArt, Runway AI, and music generator Suno.
Are any of these lawsuits succeeding?
Yes. The Verge reports that artists have won some victories, though others have dragged on for years or settled. Courts are still establishing precedent on what counts as fair use in AI training.