Jason M. Allen wanted more than a blue ribbon from the Colorado State Fair for his science-fiction-themed image “Theatre D’opera Spatial.” He wanted to own the copyright. But last fall, the U.S. Copyright Office rejected his application because he used generative AI, and his work was not the product of human authorship. 

Allen’s image is just one of many examples at the center of one of the biggest debates regarding generative AI. Some people argue that AI-generated output, including images, written content, music, and code, should not be eligible for copyright because it is not generated by a human. Others argue that Allen and those who use AI to create something should be eligible for copyright because of the human element involved in the process. 

Why Is a Copyright Important?

When someone creates something original, they want the ability to keep others from using it to make financial profits or take credit for their work. By being granted a copyright for a specific creation, which means the right to copy or use it, the original creators and their authorized agents are the only ones who can reproduce the work. Other people cannot use it without authorization; for example, a company would not be permitted to use a piece of digital art as its logo without permission. Also, copyright protects the creator from unauthorized distribution, such as a publisher producing an author’s novel without permission. 

However, copyright laws were created for humans, so the ability of machines to create using generative technology means revisiting a legal process that has existed for over 300 years. Copyrights are built on the foundational idea that when humans combine existing data, such as concepts, designs, or things, in new ways to generate new solutions, ideas, or designs, the output is usually copyrightable__ assuming no plagiarism is involved. 

The debate now centers around the results of generative AI. Supporters of copyrights for AI-generated projects argue machines use similar data (concepts, designs, or things) that humans use when creating something new. Additionally, they point out that a human still must provide the direction for the generative AI technology. 

For others, including the U.S. Copyright Office, the line in the sand is authorship. According to the decision for Allen’s copyright application, the copyright was denied because all of the work’s “traditional elements of authorship” were produced by a machine. However, suppose a copyright application is for something that contains both AI- and human-generated elements. In that case, the office may grant copyright for the human contributions if there is enough human authorship for copyright. Still, the definition of “enough human authorship” is unclear. Note that currently, the AI-generated portion will not be granted a copyright. 

Many lawsuits are currently contesting this decision not to grant copyright to AI-generated works. But even with the current ruling, the question remains: Who actually owns the copyright for AI-generated works created using data, much of which is copyrighted? The Generative Artificial Intelligence and Copyright Law, updated in September 2023, states that no clear rule has emerged identifying the “author or authors” of AI-generated work. 

Far-Reaching Impact of the AI Copyright Decisions 

On the surface, copyright decisions for AI-generated content appear to only affect creators. However, the copyright rules actually impact more than a single person and could have widespread effects. 

People are typically proud of their unique creations when they create a song, novel, image, or other output. Even when using AI to generate the output, the creator is still an integral part of the process and result. If creators cannot copyright their work generated in partnership with AI, they will likely lose their motivation because they can no longer protect their work generated using AI. 

Additionally, generative AI’s ability to continue improving output quality depends on tech providers continuing to invest in technology. Experts are concerned that without copyright, AI tech providers will have less incentive to invest in generative AI. If the capabilities do not continue to evolve, creators will begin using AI less. Therefore, the cost of creating new content will increase, resulting in a lack of accessibility and content diversification. Remember those days when we used to spend hours finding a desired image for our content or paying about $20 per image to platforms like iStock? Probably, no one is interested in going back to those days.

However, granting copyrights for AI-generated work also has consequences, especially when AI-generated content outnumbers original work. For example, if musicians can copyright AI-generated songs, then there is a concern that the majority of music in the future will be AI-generated. Even worse, this opens the scenario where tech providers could produce excessive content and flood multiple platforms with AI-generated content, putting professional creators, such as graphic designers, musicians, writers, and even other professionals, out of work. 

The Future of Copyrights and AI-Generated Content

AI-generated content has brought a new debate to the age-old institution of copyrighting. As the early cases and decisions are moving into the marketplace, we are starting to see the consequences – and often the unintended impacts – of where the U.S. Copyright Office is falling. Many experts expect the copyright decision to evolve in the coming years as both the technological capabilities and the use of generative AI continue to grow. 

What do you think? Should the results of generative AI be copyrightable?