Every AI Copyright Lawsuit Worth Knowing

The law is currently playing a game of catch-up with your hard drive. While you've been experimenting with prompts, courts have been busy deciding whether your output belongs to you or to the ether. Every ai copyright lawsuit currently moving through the system acts as a wall. These decisions will determine if your creative labor is protected property or just public domain filler. You need to know what is at stake before you hit save on your next project.

Thaler v. Perlmutter

Stephen Thaler tried to register a piece of visual art created entirely by his Creativity Machine, an AI system he built. The US Copyright Office denied him. They said human authorship is a hard requirement for protection. Thaler sued. The court sided with the Office. They ruled that human authorship is a bedrock principle of copyright law. If a human didn't guide the creative expression, the work doesn't get a certificate. You can't copyright a prompt-generated image if you didn't do the heavy lifting of composition and arrangement. This case confirms that machines are tools, not authors.

Zarya of the Dawn

This was the first time the Copyright Office walked back a registration. Kristina Kashtanova wrote a graphic novel and used Midjourney to generate the images. The Office granted a copyright for the text and the layout, but they revoked the protection for the individual AI-generated images. They argued that the images were not the product of human creative control. Kashtanova didn't choose the specific pixels. She didn't direct the machine in a way that granted her artistic ownership. It’s a warning: you own your arrangement, but you don't own the raw output of a diffusion model.

Andersen v. Stability AI

Sarah Andersen and other artists sued Stability AI, Midjourney, and DeviantArt. They claim these companies scraped billions of images without permission to train their models. The plaintiffs argue this is massive copyright infringement. The court dismissed parts of the complaint but allowed the core claims to proceed. We are watching to see if training a model on copyrighted work qualifies as fair use. Fair use is a legal defense that allows the use of protected material without permission under specific circumstances. If the court decides training is not fair use, the current AI business model is in deep trouble.

The New York Times v. OpenAI

The New York Times is suing OpenAI and Microsoft. They claim the models can reproduce near-verbatim excerpts of their journalism. This shifts the debate from training to output. If an AI regurgitates your proprietary writing, that is a direct challenge to your market. The Times wants to prove that these models are essentially high-tech plagiarism machines. This case matters because it tests whether commercial AI companies owe money to the sources they consume. If the paper wins, the industry might have to pay for the data they ingest.

Suno and Udio

The major record labels are currently suing music generators Suno and Udio. They claim these platforms are churning out sound-alikes that compete directly with human musicians. The labels are asking for statutory damages, which are set fines between $750 and $30,000 per work infringed. This is a battle over the soul of the music industry. If the labels win, the developers might be forced to delete their entire training sets. That would effectively delete the products you're using today. It's a high-stakes fight about whether an algorithm can replace a songwriter.

Why You Should Care

You aren't just a spectator. Every time you use an AI tool, you are participating in a system that is currently being stress-tested by federal judges. If you want to protect your work, you need to understand the human authorship requirement. The Copyright Office expects you to do more than just type a sentence into a box. They want to see your creative fingerprint. If you can't prove you steered the creative process, your work might end up as public domain fodder.

The filing fee for a basic claim is about $65. You usually have a 3-month window from the first publication to register for full statutory damages. If you wait longer, you lose the right to demand those specific fines and you only get actual damages, which are much harder to prove in court. Copyrightable is not a law firm. We do not provide legal advice, so you should always check the latest guidance at copyright.gov. The law is messy. It is slow. It is biased toward human effort. Keep your records, document your process, and keep your human touch visible.

You need to know if your workflow is legally sound before the next ruling drops. Install the Copyrightable hook to track how your specific creative habits align with these ongoing cases.