Copyright Office Clarifies Policies Regarding Copyright for AI Assisted Works (Sort Of)

On March 16, 2023, the U.S. Copyright Office issued a new policy directive for the registration of works created in whole or in part by artificial intelligence. 1 Unfortunately, the directive was not very clear, as it generated these headlines in response:

“AI-“Assisted” Works Are Now Open to Copyright Protection, Raising Questions for Hollywood” 2

“US Copyright Office: AI Generated Works Are Not Eligible for Copyright” 3

“U.S. Copyright Office says some AI-assisted works may be copyrighted” 4

“U.S. Copyright Office Rules A.I. Art Can’t Be Copyrighted” 5

Who’s right?

Unfortunately, both of them.

Let’s go back and look at two areas that do not have copyright protection; facts and the public domain.

Facts are ineligible for copyright protection. And, additionally, as Section 102(b) enumerates, no protection exists for any “idea, procedure, process, system, method of operation, concept, principle, or discovery.”

Yet, if the facts are arranged in a manner that shows authorship, or are explained in a manner that shows authorship, then a copyright will issue, yet the facts will remain unprotected. No history book could exist if this were not so.

Similarly, a work that is in the public domain has no copyright. But if a book consists primarily of photographs in the public domain, a copyright will issue if the photos are arranged in an original manner, such as a pictorial biography of Mark Twain, in which all of the subject photos will be in the public domain.

Such as it is with “Zarya of the Dawn,” the graphic novel which at first was granted copyright by the Copyright Office but later had its registration revoked, when it was revealed (by the author rather imprudently) that the artwork was generated by AI. 6

“Although the images cannot be copyrighted, USCO did state that [the author] can copyright the parts of the book that she wrote and how the AI-generated images are arranged on the page… USCO rescinded the original protection and issued a new copyright that only covers the writing and the layout of the images, not the images themselves. The new copyright is retroactive to the original issue date.” 7

So the new policy regards how a work which has been “assisted” in whole or in part by AI may be registered. Unfortunately, it is not the picture of clarity.

“Individuals who use AI technology in creating a work may claim copyright protection for their own contributions to that work. They must use the Standard Application, and in it identify the author(s) and provide a brief statement in the ‘Author Created’ field that describes the authorship that was contributed by a human. For example, an applicant who incorporates AI-generated text into a larger textual work should claim the portions of the textual work that is human-authored. And an applicant who creatively arranges the human and non-human content within a work should fill out the ‘Author Created’ field to claim: ‘Selection, coordination, and arrangement of [describe human-authored content] created by the author and [describe AI content] generated by artificial intelligence.’” 8


“AI-generated content that is more than de minimis should be explicitly excluded from the application. This may be done in the ‘Limitation of the Claim’ section in the ‘Other’ field, under the ‘Material Excluded’ heading. Applicants should provide a brief description of the AI-generated content, such as by entering ‘[description of content] generated by artificial intelligence.’ Applicants may also provide additional information in the ‘Note to CO’ field in the Standard Application.” 9

Now, the person who is filling out the registration form is not likely to be an attorney. It may be the authors themselves. Are they to know that the “Standard Application” must be used rather than form TX (for literary works), or form VA (for works of the visual arts) if the work contains AI material?

Plus, unless the person is familiar with the process, how are they to judge whether the contribution is de minimus and that the required information should go in the “Limitation of Claim” field rather than the “Author Created” field?

This sort of regulatory nit-picking is going to lead to various unwelcome outcomes, the most likely being that mistakes are going to be made or the author is going to say “to heck with it,” and lie about the AI contribution. None of this will matter, of course, until the copyright at issue is subject to litigation, either by the author suing for infringement or being sued for infringement. Since the AI process is the proverbial “cat let out of the bag,” an attorney, before filing suit, will need to quiz the client closely about whether AI was used, and if so, how much. Likewise, the defense attorney’s first question to the Plaintiff will be “was AI used in the creation of all or some of the work at issue?”

For the unlucky author who botches this part of the registration process, the result may be a court ruling that their copyright is invalid.

For my two cents, it would be much better for the Copyright Office to amend the registration forms to ask this point-blank question: “Was Artificial Intelligence used in whole or in part, in the creation of this work?” with a box to check “Yes” or “No.” Following up, the next box should be “If the answer is “yes,” please explain in detail the areas and extent in which AI was used.”

That would vastly clear things up. And be a lot easier for all concerned.

You can get my latest article in your email