All images that are created via text-to-image prompts are not protected under US copyright laws, the US Copyright Office (USCO) has stated. That means that anything created via Stable Diffusion, Dall-E or Midjourney cannot be copyrighted, at least in the USA.
The USCO equated the prompt-creator to telling an artist or designer what to make. That is, the idea itself is what is being discussed, and as such, the idea has no value available to be given copyright protection.
“They identify what the prompter wishes to have depicted, but the machine determines how those instructions are implemented in its output,” the USCO wrote in new guidance it published to the Federal Register.
“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,” the office stated.
The Office noted that although there was some artistic and creative human input involved in giving the prompt, the overall result was still largely dictated by the machines. At least at the moment. This is in keeping with the famous Monkey Photo copyright case from a few years ago, where it was ruled that an animal cannot hold any copyright.
The Office continued to say that “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,” the USCO said. “In the Office’s view, it is well-established that copyright can protect only material that is the product of human creativity.”
The USCO did leave a small chink open for re-evaluating the laws in future, particularly if technology changes. Judging by the early outputs from Midjourney’s latest version 5 I think we have reason to worry. The images that I’m seeing are almost indistinguishable from photographs.
I predict that it will be a very short time before brands start demanding its use for their campaigns. They won’t put up with their ad campaigns not having any copyright protection for long. I think we will start to see lobbying for change on this matter as the technology becomes more ubiquitous.
The USCO has stated that they may be open to copyrighting elements that contain AI processes. For now, they say that it will be a case-by-case process. Any traditional methods may be afforded copyright protection, even if there are other AI-generated elements involved.
The example given was the AI-generated graphic novel by Kris Kashtanova that was stripped of its copyright last year. The text and layout, however, have been granted copyright, just not the overall complete work, including the images.
Photographers and other artists who are vehemently against AI images will breathe a sigh of relief at this ruling. However, it may not be for too long, only the sands of time will tell.