The US Copyright Office recently rejected a request to let an AI copyright a work of art (applied for on the AI’s behalf by its owner, Stephen Thaler). The US District Court for DC was then asked to rule on a motion for summary judgment in an appeal of the Copyright Office decision and, on Aug 18, 2023, granted that motion in favor of the Copyright Office. (Thanks to Laercio Sousa at Velloza, the MI member firm in Brazil for highlighting the decision.)
Here are some notes:
Nothing has changed.
By denying any human involvement in the subject work, Mr. Stephen Thaler created an unusual record unlikely to be repeated at the current level of AI development by any human actually wishing to claim copyright on a work.
As noted in the District Court judgement, AI-generated works merely selected and arranged by humans are “at least partially the product of human creativity”, and thus protected by copyright. Your GPT4-assisted Harry Potter fanfic still belongs to you.
In any case, the District Court judgement was on whether the denial of copyright registration was "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law" and shouldn’t be read more expansively.
However, more questions are coming …
The District Court judgement reiterates some plain law that hints at future questions:
"human creativity is the sine qua non at the core of copyrightability".
Can “human-like” creativity suffice? If so, what will that mean?
When a camera generates a mechanical reproduction of a scene, it is copyrightable because "the photographer develops a ‘mental conception’ of the photograph, which is given its final form by that photographer’s decisions”.
Can an AI develop a "mental conception" and give it final form through a decision? Can an AI have an "intellectual invention"? Ideas? Current AI researchers are remarkably close to answering this question affirmatively.
"The act of human creation—and how to best encourage human individuals to engage in that creation and thereby promote science and the useful arts—was thus central to American copyright from its very inception. Non-human actors need no incentivization with the promise of exclusive rights under United States law, and copyright was therefore not designed to reach them."
Would an ‘incentivization test’ be required for AI ‘authorship’, if the mental conception and decision or human-like creativity questions above were definitively addressed?
… and the Copyright Office is seeking early answers.
As announced in the Federal Register, the US Copyright Office is seeking public input on AI and copyright-related concerns, including:
Copyrightability of AI-generated work without human involvement.
Addressing copyright liability in AI contexts.
AI's use of copyrighted data in training.
Potential implications of AI violating publicity rights and unfair competition laws (though not copyright issues).
Public comments began August 30th. Written comments are due on October 18th, and replies must be submitted to the Copyright Office by November 15th.