It's Alive!

May 02, 2023

TL; DR: Copyright law requires a lot of catching up to match current reality. Today’s policy appears to minimize the human creation factor related to the input prompts that drive AI outputs and maximizes the copyrightability of the manipulation of AI after-the-fact. Much like my input to Bing to “Draw me a picture of Frankenstein as AI” would likely result in the image below as being uncopyrightable, but if I then manipulated it further (not through additional creative prompts, as there is no credit for that) I would then likely be able to produce a copyright.

Back on April 23rd Canadian artist Grimes, posted on her Twitter feed:

I'll split 50% royalties on any successful AI generated song that uses my voice. Same deal as I would with any artist i collab with. Feel free to use my voice without penalty. I have no label and no legal bindings.

She included an image of a recent article depicting the deepfaking of artists Drake and The Weeknd by AI. I read this and was immediately intrigued and impressed by her business sense and the sense of inevitability she expressed. Essentially, an “If you can’t beat ‘em, join ‘em” attitude.

Almost to the person her Twitter followers were highly supportive, and she interacted with them throughout the thread when they asked questions about how she would be credited on songs generated by AI, how publishing would work and more. A very different approach from most musicians or artists in other industries, but one that I think is much more highly attuned to the realities surrounding the future of creative endeavors and AI.

In her Twitter thread user cuckercartoons seemed to represent the majority when they replied:

Whereas on the other side of the conversation was user wesblog who observed:

and user ItsMrMetaverse questioned along the same lines:

Both of the latter users’ comments completely ignoring the fact that the intent would be to use her voice patterns in combination with AI to generate songs that she did not sing to sell to her fan base that she had built up over time, taking money essentially out of her pocket were she to not receive some of the proceeds.

While the Recording Industry Association of America’s complaint last Oct that AI was infringing on the copyrights of their artists’ catalogs hasn’t been fully addressed, the US Copyright Office ruled that AI-generated art, including music, can’t be copyrighted as it is “not the product of human authorship”. This was the takeaway, but I felt the ruling really seemed to meander, and certainly left things subjectively open to interpretation.

In the opening paragraphs of the policy it states:

In February 2023, the Office concluded that a graphic novel comprised of human-authored text combined with images generated by the AI service Midjourney constituted a copyrightable work, but that the individual images themselves could not be protected by copyright.

Completely disregarding that the images, “… generated by the AI service MidJourney” were done only after extensive input and manipulation by the author and use of the AI merely as a tool to achieve her creative design. I wonder if anyone working in the Copyright office has ever actually tried to use MidJourney to create images like those found in Zara of the Dawn, the graphic novel in question? If they had they would surely understand that human creative process is a core component of achieving a consistent creative theme in multiple pieces of related artwork.

In section II. The Human Authorship Requirement the policy continues to hang its hat on this idea that these works in question occurred totally autonomously, stating:

In the Office's view, it is well-established that copyright can protect only material that is the product of human creativity. Most fundamentally, the term “author,” which is used in both the Constitution and the Copyright Act, excludes non-humans.

This is where the policy just really goes sideways for me (and I apologize for the length of the next quote, but I think it’s important to see it in its entirety):

In its leading case on authorship, the Supreme Court used language excluding non-humans in interpreting Congress's constitutional power to provide “authors” the exclusive right to their “writings.” In Burrow-Giles Lithographic Co. v. Sarony, a defendant accused of making unauthorized copies of a photograph argued that the expansion of copyright protection to photographs by Congress was unconstitutional because “a photograph is not a writing nor the production of an author” but is instead created by a camera. The Court disagreed, holding that there was “no doubt” the Constitution's Copyright Clause permitted photographs to be subject to copyright, “so far as they are representatives of original intellectual conceptions of the author.”  The Court defined an “author” as “he to whom anything owes its origin; originator; maker; one who completes a work of science or literature.”  It repeatedly referred to such “authors” as human, describing authors as a class of “persons”  and a copyright as “the exclusive right of a man to the production of his own genius or intellect.” 

I’m not suggesting here at all that it should be lawful to copy a photographer’s images. But how does using a camera as a tool to take a picture make the resulting picture copyrightable, but using an online algorithm as a tool to make a picture, video or other creative product not copyrightable? Is a camera a machine? Is a computer?

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. 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. 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.

Please tell me, beyond pointing the camera at the subject, setting the f-stop, dialing in the focus and pushing the shutter button, how is the actual creation of a photo primarily human-based by the policy definition? (Again, theoretical question based on the government’s definitions. I believe photographers are artists whose work should be protected. But I also believe that prompt engineers who interface with AI to produce creative output are also artists whose work should be protectable.) Further, “Based on the Office’s understanding…” this really answers my question about whether any of these folks have actually used MidJourney. When has any meaningful picture come out of MidJourney without 7 fingers and three elbows based on “solely a prompt” and not additional creative input by the author?

And yet, in the very next paragraph the policy seems to swing back one more time:

In other cases, however, a work containing AI-generated material will also contain sufficient human authorship to support a copyright claim. For example, a human may select or arrange AI-generated material in a sufficiently creative way that “the resulting work as a whole constitutes an original work of authorship.” Or an artist may modify material originally generated by AI technology to such a degree that the modifications meet the standard for copyright protection. In these cases, copyright will only protect the human-authored aspects of the work, which are “independent of” and do “not affect” the copyright status of the AI-generated material itself.

What they appear to be saying is that the output of AI, if further manipulated by a human, can be considered copyrightable, but the input generated by the human used to get the AI-generated output isn’t copyrightable. Much like my opening statement that if I had further edited the image of “Frankenstein as AI” in a sufficiently creative way it would be copyrightable, but my initial creative thought to input the (albeit in this example very minimal) prompt does not entitle me to copyright the output of my “creativity”. I’m sure author Mary Shelley would have a lot to say about that (obviously not comparing levels of effort here folks).

And yet, when we look back at the previous section and read:

The Court defined an “author” as “he to whom anything owes its origin; originator; maker; one who completes a work of science or literature.” 

We have to wonder, is the picture, video or other content (including my opening image) generated by AI not attributable to the coder who provided the input prompts, who originated the sequence that “made” the image? I have to think it is. I also have to think that right now our laws and regulations have a lot of catching up to do by people who actually understand the daily evolving technology.