Revisiting AI Inventorship in Thaler v. Vidal
The U.S. Court of Appeals for the Federal Circuit recently released its decision in Thaler v. Vidal, No. 21-2347 (Fed. Cir.), the much-discussed case that addresses whether artificial intelligence (AI) software can be listed as the inventor on a patent application. This post summarizes the Federal Circuit's opinion.
For a more in-depth summary of the dispute and the parties' positions, I encourage you to read this blog's earlier post covering the oral arguments in this case. The short version, however, is that the U.S. Patent and Trademark Office (USPTO) refused to allow two patent applications where an AI system was listed as the sole inventor. The USPTO concluded that both applications were incomplete because they lacked a valid inventor. According to the USPTO, the U.S. Patent Act "limit[s] inventorship to natural persons." Dr. Stephen Thaler, who filed the patents on behalf of his AI system, filed a series of appeals until the case reached the Federal Circuit.
On appeal, the Court affirmed the district court's holding that an AI could not be listed as an inventor on a patent application because the Patent Act requires that inventors must be natural persons. The Federal Circuit agreed with the USPTO's conclusion that the Patent Act expressly provides that inventors must be "individuals." Because the Patent Act itself does not define the word "individual," the Federal Circuit relied on the U.S. Supreme Court precedent in Mohamad v. Palestinian Auth., 566 U.S. 449 (2012), which explained that when used "[a]s a noun, 'individual' ordinarily means a human being, a person." The Federal Circuit also relied on definitions from the Oxford English Dictionary, Dictionary.com, the Dictionary Act and the Federal Circuit's own precedent in Univ. of Utah v. Max-Planck-Gesellschaft zur Forderung der Wissenschaften E.V., 734 F.3d 1315 (Fed. Cir. 2013), which held that "[i]nventors must be natural persons and cannot be corporations or sovereigns."
The Federal Circuit also rejected Thaler's arguments in reaching its decision:
- First, the Federal Circuit rejected Thaler's arguments that "inventor," when interpreted in the context of the Patent Act, may include non-human entities. While Thaler pointed to several provisions of the Act to support its argument, the Federal Circuit found each unpersuasive.
- Second, Thaler argued that inventions generated by AI should be patentable in order to encourage "innovation and public disclosure," but the Federal Circuit found this policy argument "speculative and lack[ing] a basis in the text of the Patent Act and in the record."
- Third, the Federal Circuit rejected Thaler's argument that allowing AI systems as inventors would support the constitutional purpose of patents "[t]o promote the progress of science and the useful arts." The Federal Circuit found this argument inapplicable because "[t]he constitutional provision he cites is a grant of legislative power to Congress; Congress has chosen to act pursuant to that power by passing the Patent Act."
- Finally, Thaler noted that South Africa has granted patents with his AI system as an inventor, but the Federal Circuit responded that the South African patent office did not interpret the U.S. Patent Act when it did so.
For now, patent inventorship remains the domain of human beings. But as AI continues to develop in the coming decades, the issue of AI inventorship is sure to resurface.