US Copyright Office Issues AI Copyright Guidance

As generative AI technologies like GPT-4 and Midjourney have quickly become more advanced and their creative use has exploded in popularity, the US Copyright Office has released guidance today to clarify when AI-generated content can be copyrighted.

The Copyright Office decided that an author couldn’t copyright individual AI images used to illustrate a comic book because each image was made by Midjourney and not by a human artist.

When deciding, the Copyright Office promised to stick to the long-standing legal rule that creative works can only be registered if real people make them. Officials have confirmed that this means AI technologies can never be considered authors.

This wasn’t the only situation that led to new rules, but it was the most recent one. The Copyright Office started an agency-wide initiative to investigate a broader range of copyright issues as AI models are used to make text, art, audio, and video change. This was partly because of the complicated questions about who wrote the comic book.

The guidance explains in more detail what can’t be protected by copyright when it comes to AI works made only by prompts and isn’t changed.

The Copyright Office compares this to giving “instructions to a commissioned artist.” A person did not make these works, so they won’t be registered.

If you want to see other recent news, you can click on the link given below:

When “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,” the guidance explains. “Based on the Office’s understanding of the generative AI technologies currently available, users do not exercise ultimate creative control over how such systems interpret prompts and generate material.”

But, just like in the Midjourney case, an author who puts generative AI images in a certain order, like designing the layout for a comic book, may be able to copyright that order if it is “sufficiently creative.”

If an author or artist has changed AI-generated work and “the modifications meet the standard for copyright protection,” the reasoning is the same. The instructions give examples like changing an AI-made image in Adobe Photoshop or changing AI-made music with guitar pedals.

But it’s clear that the Copyright Office is just starting to figure out how to handle these complicated cases, and the rules are still unclear. The guidance says that in the end, officials will decide on a case-by-case basis whether humans or machines did AI-assisted work.

“The answer will depend on the circumstances, particularly how the AI tool operates and how it was used to create the final work,” the guidance says.

Any AI-Generated Content Must Be Disclosed

Perhaps the most significant aspect of the guidance is an author’s “duty to disclose the inclusion of AI-generated content in a work submitted for registration.”

When registering works, authors must say which parts were written by humans and made by computers. If applicants don’t know how to talk about AI-generated content, the Copyright Office suggests saying that the work has an AI-generated range. That will make the office follow up with each author to help them fill out their applications.

The Copyright Office suggests that artists with pending applications or who have already registered works that have content made by AI fix the public record by submitting an additional registration.

The office warned that if the role of AI in copyrighted works is not shown accurately, they “lose the benefits of the registration.” That could make it easy for people to copy jobs, and there might be little or no legal recourse for claims of copying.

Failure to disclose AI-generated content is the only type of infringement discussed in the guidance. Critics like Alex J. Champandard, a co-founder of Creative. AI—a group of hackers and artists interested in generative AI—tweeted that current guidance puts authors in a precarious catch-22 situation.

“By disclosing the AI, you’re opening yourself up to infringement, but by not disclosing AI, it’s safer but in violation of [the US Copyright Office]!” Champandard’s tweet suggested.

Ars couldn’t contact Champion to discuss his group’s worries about the guidance.

More guidance is coming in 2023

The Copyright Office knows it needs to do a lot more work to make it clear when content that was made with AI can be registered. It has set up many public meetings in April and May to hear what people say.

The office also launched a webpage to post updates on AI news and events. This will make it easier to keep up with rules as they change.

On April 19, the listening sessions start with an event about literary works made by AI. The next session, on May 2, is about visual arts. On May 17, it’s about audiovisual works; on May 31, it’s about music and sound recordings.

Artists, people who work in the creative industry, people who develop AI, people who study AI, and lawyers are all encouraged to sign up to attend these sessions.

But because there will be a limit on how many people can hear, the Copyright Office said in today’s guidance that there would be other ways for stakeholders to have their say besides attending these sessions.

“The Office intends to publish a notice of inquiry later this year seeking public input on additional legal and policy topics, including how the law should apply to the use of copyrighted works in AI training and the resulting treatment of outputs,” the guidance says.

It will likely be a daunting task evolving guidance, but Nora Scheland, a public affairs specialist for the US Copyright Office, told Ars that the office is “excited” about launching a wider initiative exploring AI and will “look forward to public input and feedback throughout the coming months.”

Leave a Comment