AI Search
AI-generated content isn’t protected by U.S. copyright laws. But there are still a lot of legal questions to untangle.
At the moment, works created solely by artificial intelligence — even if produced from a text prompt written by a human — are not protected by copyright in the United States.
When it comes to training AI models, however, the use of copyrighted materials falls into a legal gray area, especially if the outputs compete with original works. For now, the fair use doctrine permits the use of copyrighted material under certain conditions without needing the permission of the owner. But pending lawsuits could change this.
AI-generated content refers to written text, video, code, audio and other media produced by generative AI tools. These machines are trained on large amounts of data, allowing them to create relevant outputs in response to a word, phrase, question or other kind of input.
Generative AI has significantly altered the way we live, work and create in a short amount of time. As a result, the deluge of AI-generated text, images and music — and the process used to create them — has prompted a series of complicated legal questions. And they are challenging our understanding of ownership, fairness and the very nature of creativity itself.
Related Reading4 Ethical Questions Generative AI Poses
It has long been the posture of the U.S. Copyright Office that there is no copyright protection for works created by non-humans, including machines. Therefore, the product of a generative AI model cannot be copyrighted.
The root of this issue lies in the way generative AI systems are trained. Like most other machine learning models, they work by identifying and replicating patterns in data. So, in order to generate an output like a written sentence or picture, it must first learn from the real work of actual humans.
If an AI image generator produces art that resembles the work of Georgia O’Keefe, for example, that means it may have been trained on works that closely resemble or include the actual art of Georgia O’Keefe, depending on the dataset used. Similarly, for an AI content generator to write in the style of Toni Morrison, it was likely trained with words written by Toni Morrison.
Legally, these AI systems — including image generators, AI music generators and chatbots like ChatGPT — cannot be considered the author of the material they produce. Their outputs are simply a culmination of human-made work, much of which has been scraped from the internet and is copyright-protected in one way or another.
So, how do we reconcile the rapidly evolving artificial intelligence industry with the knotty particulars of U.S. copyright law? That is something creatives, companies, courts and the United States government are trying to figure out.
Related ReadingThe Unreal Slim Shady: How We Trained an AI to Simulate Eminem’s Style
Creative work that is the result of a collaboration between a human and machine, which is often the case with AI-generated creations, is a complicated matter.
“If a machine and a human work together, but you can separate what each of them has done, then [copyright] will only focus on the human part,” Daniel Gervais, a professor at Vanderbilt Law School, told Built In. He mainly focuses on intellectual property law, and has written extensively on how it relates to artificial intelligence.
If the human and machine’s contributions are more intertwined, a work’s eligibility for copyright depends on how much control or influence the human author had on the machine’s outputs. “It really needs to be an authorial kind of contribution,” Gervais said. “In that case, the fact that you worked with a machine would not exclude copyright protection.”
This threshold was put to the test in September of 2022, when the U.S. Copyright Office made history by granting the first known registration of a work produced with the help of text-to-image generator Midjourney: a graphic novel called Zarya of the Dawn. Written by Kristina Kashtanova, the 18-page narrative had all the trappings of a typical comic book — characters, dialogue and plenty of images, all of which were generated using Midjourney.
Just a few months later, the office reconsidered its decision and wound up partially canceling the work’s copyright registration, claiming in a letter to Kashtanova’s attorney that it had “non-human authorship” that had not been taken into account. The book’s text, as well as the “selection, coordination, and arrangement” of its “written and visual elements,” remained protected. The images themselves did not, though, because they were “not the product of human authorship,” but rather of text prompts that generated unpredictable outputs based on its training data. The office also deemed whatever editing Kashtanova did to the images as “too minor and imperceptible to supply the necessary creativity for copyright protection.”
The office has since released a more sweeping policy change to address all AI-human creative collaborations moving forward — a response to what it sees as new trends in registration activity. The document essentially doubles down on its stance with Zarya of the Dawn, reiterating that the term “author” is not extended to non-humans, including machines. It also states that if a human simply types in a prompt and a machine generates complex written, visual or musical works in response, the “traditional elements of authorship” have been executed by AI, a non-human. Therefore, it is not protected by copyright.
Federal courts have also affirmed the U.S. Copyright Office’s position that AI-created artwork cannot be copyrighted. In August 2023, a judge in the U.S. District Court for the District of Columbia sided with the agency against computer scientist Stephen Thaler, who was seeking copyright protection for an image created by AI software.
Some creators and companies believe their content has been stolen by generative AI companies, and are now seeking to strip these companies of the protective shield of fair use in a series of pending lawsuits.
Getty Images is suing Stability AI (the company behind Stable Diffusion) for copying and processing millions of images that are protected by copyright, as well as their associated metadata owned by Getty Images, without getting permission or providing compensation.
TikTok settled a lawsuit in 2021 with voice actress Bev Standing, who claims the company used her voice without permission for its text-to-speech feature.
The New York Times joined the legal struggle as well in 2023 and again in 2025, first suing Open AI and Microsoft in 2023 for using millions of NYT articles without compensation in order to train AI models, and then suing Perplexity in 2025 for scraping NYT content and providing information that directly competes with what the publication offers readers. According to NPR’s sources, if courts find that OpenAI illegally used Times articles to train its models, OpenAI could be forced to destroy its LLM data set and rebuild it from scratch. Additional U.S. newspapers, including The New York Daily News and the Chicago Tribune, have sued Microsoft and Open AI, creating more legal headaches for the AI duo.
Major music publishers like Universal Music are targeting Anthropic, claiming the company illegally trained its chatbot Claude AI on copyrighted song lyrics. The publishers also called for greater “guardrails” to ensure Anthropic doesn’t copy song lyrics moving forward. Meanwhile, AI music startups Suno and Uncharted Labs are the center of another AI music lawsuit, led by record labels like Sony Music Entertainment and Warner Music Group.
In June 2025, one of these lawsuits reached its conclusion, with a federal judge in San Francisco handing a mixed ruling to a group of authors who alleged that Anthropic unfairly used their published works to train its models. The judge held that AI companies may legally use copyrighted materials to train their large language models — so long as they obtain the works legally — but found that the manner in which Anthropic acquired some of its training data constituted piracy. The company agreed to pay $1.5 billion to settle the lawsuit (about $3,000 per pirated work), making it the largest payout in the history of U.S. copyright cases.
Artists Sarah Anderson, Kelly McKernan and Karla Ortiz have filed a class-action copyright infringement lawsuit against both Stability AI and Midjourney, both of which use Stable Diffusion to generate their images. The suit claims that these artists’ work was wrongfully used to train Stable Diffusion, and that the images generated in the style of those authors directly compete with their own work — an important point in the matter of fair use.
“Until now, when a purchaser seeks a new image ‘in the style’ of a given artist, they must pay to commission or license an original image from that artist. Now, those purchasers can use the artist’s works contained in Stable Diffusion along with the artist’s name to generate new works in the artist’s style without compensating the artist at all,” the complaint reads. “The harm to artists is not hypothetical — works generated by AI image products ‘in the style’ of a particular artist are already sold on the internet, siphoning commissions from the artists themselves.”
In August 2024, U.S. District Judge William Orrick upheld all copyright infringement and trademark claims in the case. The lawsuit is set to move forward, providing an early victory for artists while dealing a blow to any AI company that uses Stable Diffusion.
Writers have also tried to bring class-action lawsuits against top AI companies. Nonfiction writers Nicholas Basbanes and Nicholas Gage have sued OpenAI and Microsoft, claiming the companies “simply stole” content from their works and must compensate them. The pair of writers want to represent a class of writers that could number in the tens of thousands. This comes on the heels of fiction writers suing OpenAI and seeking to establish a class-action lawsuit in late 2023.
The U.S. Copyright Office’s stance on excluding machines from being considered authors could throw a wrench in the Stable Diffusion lawsuit and many others, according to Rob Heverly, an associate professor at Albany Law School who specializes in the intersection of technology and law.
“In order for there to be infringement, there has to be an author. So, if there isn’t an author, I don’t know that there can be infringement,” Heverly told Built In. “If we’re not going to hold the technology maker liable for the technology itself, then the creator of the output is the AI. But we’ve already said they’re not an author. So if they’re not an author then they can’t create an infringing work.”
Amid all these lawsuits, the U.S. Copyright Office concluded in May 2025 that AI developers who use copyrighted works to train models that generate “expressive content that competes with” original works are going beyond the scope of the fair use doctrine. According to the report, fair use does not apply when AI outputs closely resemble and compete with original works in their existing markets. For example, if a model trained on copyrighted horror novels is used to generate a book that mimics the style and themes of a specific horror author, the resulting content would directly compete with the author’s work.
Still, the fair use doctrine’s place in the ongoing legal saga of generative AI is up in the air as the courts continue to untangle a web of lawsuits and public discussions on AI and fair use persist.
For instance, in September 2025, technology policy coalition group Chamber of Progress used the Trump administration’s request for inputs on AI regulation to argue that the use of copyrighted material for AI model training should be classified as fair use, exempting companies from being liable of statutory damages for the practice. Furthermore, they urged the U.S. government to intervene by seeking to rationalize and reform the framework of high statutory damages, arguing that the threat of lawsuits can hinder AI innovation and coerce AI companies into settlements.
Related ReadingIs Generative AI the Next Tech Bubble?
While generative AI hangs in legal limbo, creators are still worried about their work or style being used to train generators without permission or compensation.
“The large majority of independent artists make their living through commissioned works,” Ben Zhao, a computer science professor at the University of Chicago, told Built In. But, the websites they post their work on are being scraped by AI models in order to learn and then mimic that particular style. “Artists are literally being replaced by models that have been trained on their own work.”
To help, Zhao and his team designed a new tool called Glaze. If an artist wants to put a creation online without the threat of an image generator copying their style, they can upload it to Glaze first and choose an art style different from their own. The software then makes mathematical changes to the artist’s work on a pixel level so it looks different to a computer, rendering it useless as an effective piece of training data.
Other companies are taking a more offensive approach. Shutterstock, a stock imagery site that was “critical” to the training of OpenAI’s DALL-E, according to CEO Sam Altman, has gone so far as to pay content creators if their work is used to develop generative AI models.
And Shutterstock isn’t alone. Generative AI startup Bria trains its models exclusively on what it calls “responsibly sourced” data sets, and it pays royalties to artists and stock image providers when their creations have been used to generate an image. “We pay back a royalty according to the output,” co-founder and CEO Yair Adato explained. “So if somebody generates a specific art in the style of the artist, then the artist will have the right to say how much money he wants on this synthetic creation. And then we will split the revenue.”
If the use of creators’ work in generative AI models continues to go unchecked, many experts in this space believe it could spell big trouble — not only for the human creators themselves, but the technology too.
“When these AI models start to hurt the very people who generate the data that it feeds on — the artists — it’s destroying its own future,” Zhao said. “So really, when you think about it, it is in the best interest of AI models and model creators to help preserve these industries. So that there is a sustainable cycle of creativity and improvement for the models.”
In the United States, much of this preservation will be incumbent on the courts, where creators and companies are duking it out right now. Looking ahead, the level at which U.S. courts protect and measure human-made inputs in generative AI models could be reminiscent of what we’ve seen globally, particularly in other Western nations.
The United Kingdom is one of only a handful of countries to offer copyright protection for works generated solely by a computer. The European Union, which has a much more preemptive approach to legislation than the U.S., has crafted a sweeping AI Act that has taken effect and is set to address a lot of the concerns with generative AI.
If it is ultimately determined that AI companies have infringed on certain creators’ copyrighted work, it could mean a lot more lawsuits in the coming years — and a potentially expensive penalty for the companies at fault.
“One thing you have to know about copyright law is, for infringement of one thing only — it could be a text, an image, a song — you can ask the court for $150,000,” Gervais said. “So imagine the people who are scraping millions and millions of works.”
Court rulings, legal settlements and legislative proposals are rapidly reshaping the boundaries between copyright protection and artificial intelligence.
Below are some of the major developments that illustrate how policymakers, judges and industry groups are responding to disputes over creative ownership and AI innovation.
Bandcamp, a major music distribution platform, announced a policy prohibiting music that has been generated “wholly or in substantial part” by AI in order to protect human creativity and fan confidence on its platform. The company also prohibited using AI tools to impersonate other artists or styles, and clarified that data scraping or training AI models on its content is strictly forbidden. To enforce these rules, Bandcamp has encouraged users to flag suspicious content through its reporting tools and reserves the right to remove any music tracks suspected of being AI-generated.
Disney and OpenAI struck a $1 billion, three-year partnership, marking a major turning point in the AI copyright battle. Under the agreement, Disney will license more than 200 characters from its Disney, Marvel, Pixar and Star Wars properties for use in OpenAI’s Sora AI video generator, allowing users to create short, prompt-driven videos featuring icons like Mickey Mouse and Darth Vader beginning in early 2026. Disney will also become a major OpenAI customer—using ChatGPT and OpenAI APIs to build internal tools and new offerings for Disney+—and receive equity plus warrants as part of its investment. The deal signals a strategic shift from legal opposition to collaboration between Hollywood and the generative AI industry, giving Disney influence over how its intellectual property is used while also setting a potential blueprint for AI licensing in entertainment.
On December 5, 2025, The New York Times filed a federal lawsuit against Perplexity AI, alleging the AI search engine illegally copied and distributed millions of its copyrighted articles, videos and podcasts. The complaint claims Perplexity’s “answer engine” generates responses that are often “verbatim or substantially similar” to Times content, effectively creating a substitute product that deprives the publisher of subscription and advertising revenue. Furthermore, The New York Times accused Perplexity of trademark dilution for generating AI hallucinations — fabricated content falsely attributed to the newspaper.
On December 4, 2025, the Chicago Tribune filed a copyright infringement lawsuit against AI search startup Perplexity AI, accusing the company of unlawfully scraping millions of its copyrighted articles. The suit alleges that the Perplexity engine systematically copies and distributes the Chicago Tribune’s content to generate direct, substitutive answers, thereby bypassing the newspaper’s paywalls and stealing subscription and advertising revenue. This legal action is a significant point in the growing conflict over how AI companies collect and repurpose copyrighted journalistic material.
A regional court in Munich, Germany reportedly ruled that OpenAI violated the nation’s copyright laws by training ChatGPT’s underlying models on licensed musical work without permission. The decision stems from a November 2024 lawsuit filed by GEMA, the group that handles music rights in Germany, who claims the company scraped protected lyrics by popular artists to hone the chatbot’s language capabilities.
OpenAI has been ordered to pay an undisclosed amount in damages to GEMA, but said in a statement that it “disagrees with the ruling“ and is “considering next steps.” Meanwhile, GEMA celebrated the decision as “the first landmark AI ruling in Europe,” saying it could have implications for other types of art as well.
“Today, we have set a precedent that protects and clarifies the rights of authors: even operators of AI tools such as ChatGPT must comply with copyright law,” GEMA chief executive Tobias Holzmüller said in a statement. “Today, we have successfully defended the livelihoods of music creators.”
U.S. Southern District of New York Judge Sidney Stein allowed a class-action lawsuit brought by numerous authors, including George R.R. Martin, against OpenAI and Microsoft to proceed. The judge ruled that a reasonable jury could find that the allegedly infringing outputs generated by AI, such as an AI-generated outline for a Game of Thrones sequel, to be “substantially similar” to the plaintiffs’ copyrighted works, thereby potentially constituting copyright infringement.
While the decision allows the authors’ claims regarding infringing AI outputs to move forward, it did not resolve whether the defendants are protected by the key legal defense of “fair use.”
Anthropic agreed to a landmark $1.5 billion class-action settlement applying to approximately 500,000 works, compensating around $3,000 per work and requiring the destruction of improperly acquired content — a landmark resolution in AI copyright litigation from the Bartz v. Anthropic case. This settlement is the largest public copyright recovery in U.S. history and may catalyze licensing frameworks in the AI industry.
In July 2025, in the Northern District of California, District Judge William Alsup ruled that using legally purchased copyrighted content for training large language models (in the case Bartz v. Anthropic) constituted fair use. These outcomes contrast with earlier rulings, demonstrating emerging judicial nuances in balancing transformative use against infringement.
On May 9, 2025, the U.S. Copyright Office released the pre-publication version of Part 3 of its Copyright and Artificial Intelligence report, focusing on generative AI training. It concluded that using copyrighted materials for AI model development may constitute prima facie infringement, warned that models themselves could infringe if outputs closely resemble training data and emphasized that “transformative” arguments are not inherently valid.
In Thomson Reuters v. Ross Intelligence, the U.S. District Court for Delaware ruled that ROSS Intelligence using copyrighted Westlaw headnotes to train an AI-driven legal research search tool did not qualify for fair use — marking a foundational precedent rejecting AI training under the doctrine in certain contexts.
In April 2024, California Representative Adam Schiff introduced the Generative AI Copyright Disclosure Act to the U.S. Congress, requiring AI firms to notify the Register of Copyrights at the U.S. Copyright Office of copyrighted works used in generative AI training at least 30 days before publicly releasing an AI model. The bill was not signed into law, but it was supported by creative industry associations and reflected growing pressure for transparency in AI data usage.
On March 21, 2024, the ELVIS Act (Ensuring Likeness Voice and Image Security Act) was signed into law in Tennessee by Governor Bill Lee, which criminalizes the unauthorized cloning of performer, songwriter and music industry professional voices using AI technologies. The ELVIS Act is the first enacted U.S. legislation — and Tennessee is the first U.S. state — to specifically protect musicians from unauthorized use of their voices via AI. Effective July 1, 2024, the ELVIS Act signaled early legislative engagement with AI-driven intellectual property risks.
No, AI content and any works created solely by AI cannot be copyrighted in the United States.
It depends — generative AI may violate copyright laws when the program has access to a copyright owner’s works and is generating outputs that are “substantially similar” to the copyright owner’s existing works, according to the U.S. Congressional Research Service. However, there is no federal legal consensus for determining substantial similarity.
Training generative AI models using copyrighted materials is protected under certain conditions by the fair use doctrine of the U.S. copyright statute.
Companies that have developed and are responsible for AI systems may be able to be sued for copyright infringement. There are several cases of AI companies being sued due to potentially using copyrighted works to illegally train AI models or generate AI content.
For a product to be copyrighted, a human creator is needed. AI-generated content can’t be copyrighted because it isn’t considered to be the work of a human creator.
No, fully AI-generated art cannot be copyrighted in the United States. Just like any other type of AI-generated content, AI art isn’t considered to be the work of a human creator. Because AI isn’t legally viewed as an author either, no author can copyright fully AI-generated art.
When using generative AI tools, review any terms of service, license agreements or contracts. These clarify what a tool’s intended purpose is and whether any content created with it can be used for commercial purposes.
This content is for informational and educational purposes only. Built In strives to maintain accuracy in all its editorial coverage, but it is not intended to be a substitute for financial or legal advice.
Andreas Rekdal and Matthew Urwin contributed reporting to this story.
