In today’s digital age, the concept of copyright has become more relevant than ever, especially as generative artificial intelligence (AI) technologies rapidly evolve. While many people might not frequently consider copyright issues, nearly everyone is a copyright owner or author in some form—whether through writing, art, music, or other creative endeavors. However, the intersection of copyright and AI has become a complex, often contentious topic, raising critical questions about ownership, legality, and the future of creativity.
Generative AI tools, such as chatbots, image creators, and video generators, rely heavily on vast amounts of high-quality, human-generated content to train their models. This content helps these AI systems develop more lifelike personalities, artistic styles, and creative outputs. At the same time, there is growing curiosity and debate about whether AI-generated works themselves can qualify for copyright protection. These dual issues—how AI uses existing copyrighted materials and whether AI-produced creations can be copyrighted—are reshaping the landscape of intellectual property law.
Unfortunately, the current state of copyright law as it relates to AI is far from settled. Many AI companies have been vague or secretive about the specific datasets they use to train their models, leading to a flood of legal challenges. More than 30 lawsuits are currently underway in the United States involving AI companies accused of using copyrighted materials without permission. One prominent case involves The New York Times suing OpenAI, alleging that ChatGPT incorporated reporters’ articles verbatim without authorization or credit. Similarly, CNET’s parent company, Ziff Davis, filed a lawsuit against OpenAI in April, claiming copyright infringement related to training and operating AI systems.
For those covering or involved in AI and copyright issues, these developments have created an urgent need to understand the nuances of copyright law as applied to AI. Copyright, under the US Copyright Act of 1976, protects “original works of authorship fixed in any tangible medium of expression,” including books, art, music, films, software, blogs, choreography, and more. As the US Copyright Office succinctly puts it, once you create an original work and fix it in some form—whether it’s a photograph, poem, or song—you are automatically the author and copyright owner.
There are two main ways copyright and AI intersect. First, on the output side, users of AI tools want to know if the works they create with AI—be it text, images, or videos—can be protected by copyright. Second, on the input or development side, there are serious concerns about AI companies potentially using copyrighted materials illegally to train their models.
Regarding AI-generated outputs, the US Copyright Office has been clear in its guidance: works created entirely by AI without human authorship are generally not eligible for copyright protection. However, many AI tools today function more as editing or augmenting assistants rather than fully autonomous creators. For example, AI can be used to de-age actors in videos, remove objects from images, or refine audio tracks. In these cases, where human creators contribute significant creative input alongside AI, copyright protection may still apply, provided the AI’s role is properly disclosed during registration.
There are rare exceptions where fully AI-generated works have received copyright protection, but these cases require proof that a human’s creative input or manipulation of the AI-generated elements was substantial enough to warrant authorship rights. One company notably succeeded in this, demonstrating how human creativity intertwined with AI output can cross the threshold for copyright.
On the other hand, copyright law is fundamentally designed to give rights holders control over how their works are used. Most creators license their work, allowing others to use it legally under specific conditions, often involving fees and attribution. Some publishers and media companies have proactively made deals with AI firms, granting permission and licensing their content for AI training. The Financial Times and Axel Springer brands, for instance, have struck multimillion-dollar agreements to this effect.
However, many AI companies have reportedly used copyrighted content without such permissions, sparking legal battles. A notable example is a class-action lawsuit by concept artist Karla Ortiz against Stability AI. The core legal question is whether these companies’ use of copyrighted works in AI training constitutes infringement, which involves reproducing, distributing, or creating derivative works without authorization.
Decades of copyright law precedent generally prohibit unauthorized use, but the courts will ultimately determine how these principles apply to AI development. Meanwhile, AI companies are advocating for a “fair use” exception—a legal doctrine that allows limited use of copyrighted material without permission under certain