The US Copyright Office will consider an AI-generated work copyrightable if a human can prove they themselves put a meaningful amount of creative effort into the final content, according to a policy published on Thursday.
AI software capable of automatically generating images or text from an input prompt or instruction has made it easier for people to churn out content. Correspondingly, the USCO has received an increasing number of applications to register copyright protections for material, especially artwork, created using such tools.
US law states that intellectual property can be copyrighted only if it was the product of human creativity, and the USCO only acknowledges work authored by humans at present. Machines and generative AI algorithms, therefore, cannot be authors, and their outputs are not copyrightable.
Digital art, poems, and books generated using tools like DALL-E, Stable Diffusion, Midjourney, ChatGPT, or even the newly released GPT-4 will not be protected by copyright if they were created by humans using only a text description or prompt, USCO director Shira Perlmutter warned.
“If a work’s traditional elements of authorship were produced by a machine, the work lacks human authorship and the Office will not register it,” she wrote in a document outlining copyright guidelines.
“For example, when an AI technology receives solely a prompt from a human and produces complex written, visual, or musical works in response, the ‘traditional elements of authorship’ are determined and executed by the technology – not the human user.
“Instead, these prompts function more like instructions to a commissioned artist – they identify what the prompter wishes to have depicted, but the machine determines how those instructions are implemented in its output.”
The USCO will consider content created using AI if a human author has crafted something beyond the machine’s direct output. A digital artwork that was formed from a prompt, and then edited further using Photoshop, for example, is more likely to be accepted by the office. The initial image created using AI would not be copyrightable, but the final product produced by the artist might be.
Thus it would appear the USCO is simply saying: yes, if you use an AI-powered application to help create something, you have a reasonable chance at applying for copyright, just as if you used non-AI software. If it’s purely machine-made from a prompt, you need to put some more human effort into it.
In a recent case, officials registered a copyright certificate for a graphic novel containing images created using Midjourney. The overall composition and words were protected by copyright since they were selected and arranged by a human, but the individual images themselves were not.
“In the case of works containing AI-generated material, the Office will consider whether the AI contributions are the result of ‘mechanical reproduction’ or instead of an author’s ‘own original mental conception, to which [the author] gave visible form’. The answer will depend on the circumstances, particularly how the AI tool operates and how it was used to create the final work. This is necessarily a case-by-case inquiry,” the USCO declared.
Perlmutter urged people applying for copyright protection for any material generated using AI to state clearly how the software was used to create the content, and show which parts of the work were created by humans. If they fail to disclose this information accurately, or try to hide the fact it was generated by AI, USCO will cancel their certificate of registration and their work may not be protected by copyright law. ®
