SCOTUS refuses Thaler’s appeal to patent AI inventions
Supreme Court dismisses scientist’s challenge to lower court’s ruling | Decision marks the end of the road for Thaler’s DABUS campaign in the US—at least for now.
Scientist Stephen Thaler has lost his long-running battle in the US to name his artificial intelligence (AI) machine as an inventor, after the US Supreme Court (SCOTUS) denied his appeal for a writ of certiorari.
The decision, delivered yesterday, April 24, effectively draws a line under Thaler’s campaign to name an AI as a sole inventor on a patent, confirming the requirement by the US Patent Act for an inventor to be human.
After multiple rejections of his patent applications, Thaler submitted a petition to SCOTUS on March 17, challenging an appeals court ruling last August that patents can only be granted to human inventors, not machines such as Thaler’s DABUS.
The central question in Thaler’s petition was: “Does the Patent Act categorically restrict the statutory term ‘inventor’ to human beings alone?”
Key points, among others, made in the petition stated that the case “presents an ideal vehicle for resolving the question presented” and that “the patentability of AI-generated inventions is exceptionally important to the United States and across the world”.
Thaler’s appeal was supported by academics and lawyers in various amicus curiae briefs.
The US Court of Appeals for the Federal Circuit had supported a decision by the US Patent and Trademark Office ( USPTO) in May 2020 to reject the physicist’s patent applications on the two inventions on the basis that the inventor was an AI system.
As well as rebuttals from the US, Thaler has already had his DABUS-related patent applications rejected by the UK, the European Patent Office, and Australia.
South Africa is so far the only jurisdiction to have granted the patents, in July 2021—the first patent office in the world to grant a patent listing an AI rather than a person as the inventor. However, South Africa does not examine patent applications in a substantive way.
AI creations
Thaler, who masterminded DABUS in 2019, has attempted relentlessly to patent two inventions that he says were created by the AI machine without human input.
The inventions are a flexible food and beverage storage container based on fractal geometry (basically, interlocking food containers that are easy for robots to grasp and stack). The other is an emergency LED light beacon related to the container, which has a lacunar pulse train to generate a ‘neural flame’, and acts as a beacon for human attention.
Thaler’s petition in March stated: “...it is undisputed that an AI system known as DABUS used generalised background knowledge of a technical field to conceive of two novel inventions and then recognise their utility, all without specific guidance from a human being.
“Thus, only DABUS fits the statutory definition of ‘inventor’ under the Patent Act: the ‘individual…who invented or discovered the subject matter of the invention.’”
Concerns
Dennis Crouch, law professor at the University of Missouri School of Law, expressed his misgivings following the decision in his blog Patently-O.
“I am quite concerned for the role of patent attorneys and the upcoming ethical dilemmas—that patent attorneys will be prompted to bury the truth about AI contributions within their patent applications.”
He added that the situation calls for guidance from the USPTO or the legislature on the definition of “inventor” in the context of AI-generated inventions, since “the current legal framework does not adequately address this evolving landscape of innovation driven by AI”.”
‘Change will come’
Ryan Meyer, an attorney at Dorsey & Whitney in Seattle, said of the decision: “For now, AIs are barred from sole inventorship until Congress changes the law. However, AI’s rapid development and growing significance guarantees that change will eventually come.”
He added that while AI creators, tech companies, and other users can still use AI to create inventions, those inventions won’t be protected by the limited monopoly conferred by patents.
“As an alternative, AI users should consider whether they can use an AI-generated invention as a confidential trade secret, which doesn’t have the public disclosure and inventorship requirements of a patent,” he suggested.
“Courts and the law have not yet squarely addressed whether an invention conceived jointly by a human and an AI can be patented. Such a human-machine partnership would be consistent with the general rule that contributing to even part of a single patent claim can be enough to qualify as a co-inventor.
“A patent application with human and AI co-inventors will likely be the next test of the law.”
Copyright clarification
The decision comes in the wake of guidance published by the US Copyright Office ( USCO) in an attempt to clarify copyright concerns related to works of art generated, in full or in part, by artificial intelligence (AI).
The guidance was likely prompted by Thaler’s case, as well the retraction of copyright by the USCO for graphic novelist Kris Kashtanova’s graphic novel, which used an artificial intelligence (AI) tool to generate the book’s images.
In a statement issued on March 15, the office decreed that under certain circumstances, some works of art created with the assistance of AI may be copyrighted.
Commenting on this guidance, Meyer suggested that the USPTO “might issue analogous guidance in the next several months”.
“The drafters of our patent and copyright laws likely never contemplated machines that could invent and create works of art on their own, which could be why courts and government agencies have consistently interpreted words like ‘inventor,’ ‘individual,’ and ‘author’ to refer only to a human,” he added.
“Although Thaler’s appeals were arguably a longshot, his ultimate defeat was not a foregone conclusion given the ambiguity of those words.”
Stakeholder input on AI
The Supreme Court’s rejection of Thaler’s appeal coincides with an AI Inventorship Listening Session organised by the USPTO, in a bid to seek stakeholder input on the “current state of AI technologies and inventorship issues that may arise in view of the advancement of these technologies”.
The session follows a Federal Register Notice (FRN) published on February 14 requesting comments on the subjects.
There will be a second AI inventorship listening session on May 8, as outlined in an FRN issued on April 17.
Did you enjoy reading this story? Sign up to our free daily newsletters and get stories sent like this straight to your inbox.
Today’s top stories
