The U.S. Supreme Court has refused to hear an appeal in Thaler v. Perlmutter, a prominent copyright dispute over whether works generated solely by artificial intelligence can receive copyright protection under U.S. law. The Court’s denial leaves in place an earlier ruling from the U.S. Court of Appeals for the D.C. Circuit affirming that copyrightable works must have identifiable human authorship, and that an AI system acting alone cannot qualify as an “author” for copyright purposes. This effectively maintains the current human-authorship rule under U.S. copyright practice. See the Thaler v. Perlmutter opinion here: https://law.justia.com/cases/federal/appellate-courts/cadc/23-5233/23-5233-2025-03-18.html.
Key points from the U.S. case:
- Dr. Stephen Thaler’s generative AI system produced a work independently, and he sought a copyright by naming the system as the author.
- Both the Copyright Office and the federal courts concluded that works must be authored by a human to be eligible for registration.
- Issues about how much human influence might suffice in mixed human/AI creations remain unresolved.
In Europe, the Munich Local Court (Amtsgericht München) issued a decision on February 13, 2026 (Case No. 142 C 9786/25) addressing the copyright status of AI-generated logos. In its ruling, the court held that three AI-generated logo designs were not eligible for copyright protection under German law because the human prompting did not rise to the level of a creative, personal contribution that would make the logos “works” reflecting an author’s individuality. The full German judgment is available here: https://www.gesetze-bayern.de/Content/Document/Y-300-Z-BECKRS-B-2026-N-1513?hl=true.
Highlights from the Munich decision:
- The decisive question is whether the final output objectively reflects the human creator’s personality and free creative choices.
- The court examined different prompt scenarios, including very short prompts and highly detailed instructions.
- In this case, even an extensive prompting (including a 1,700-character instruction) was insufficient where the human input did not demonstrably determine the expressive elements of the final design.
- The court stressed that merely initiating, refining, or technically adjusting AI output does not automatically amount to copyright-relevant creative authorship.
Short comparison:
These parallel developments in the U.S. and Germany illustrate a common — and increasingly influential — theme in the legal treatment of AI output: copyright protection remains fundamentally tied to meaningful human creative contribution, and where AI operates with minimal personally attributable human creative input, the resulting output may not qualify for protection. While both U.S. and German courts ultimately denied copyright protection for purely AI-generated works, their reasoning reflects different legal frameworks and emphases.
The U.S. courts focused primarily on a statutory interpretation question: the US Copyright Act presupposes a human author, and an AI system cannot legally qualify as such. The analysis is therefore categorical — without human authorship, protection fails at the threshold. The Munich court, by contrast, interpreting its own law, did not frame the issue as whether AI can be an “author,” but instead conducted a fact-specific originality analysis under German copyright doctrine. The court examined whether the human’s prompting and interaction with the AI resulted in a work that reflects the author’s personal intellectual creation. The German approach is thus less categorical and more centered on whether sufficient human creative influence can be demonstrated in the specific case. However, it requires a case-by case analysis of the human input (in this case for creating a logo) - even if the author uses a lengthy prompt.
