By Dr. Raj Davé and Om Desai
US District Judge Leonie Brinkema concludes DABUS, a robot created by Dr. Stephen Thaler, cannot be an inventor.
A typical science fiction trope: robots become more intelligent than men and take over the world.
But what used to be science fiction is quickly becoming a reality. On May 14, of 2020, the first installment of this article was published discussing the original rejection of an AI-invented patent.
The story begins with University of Surrey Law Professor, Ryan Abbott, who leads the Artificial Inventor Project. He enlisted the help of the founder of Imagination Engines Inc, Dr. Stephen Thaler, to create DABUS, artificial intelligence (AI) specifically designed to invent.
DABUS then created an innovative beverage holder and a flashing light device. Dr. Thaler’s next step was to apply for patents on these two items in multiple countries. Australia and South Africa both ruled in favor of DABUS, officially labeling the AI as the inventor and Dr. Thaler as the patent owner.
The UK and Europe both rejected the patent on the basis that AI was not a person who could apply for a patent. Then, the contest came to the US. The application had been filed in 2018 and was rejected two years later, in April of 2020. The reasoning cited was that the inventor was not a “natural person.”
In a previous article titled “Should a robot be an inventor?”, Dr. Raj Dave and Mr. Krishna Shastri mentioned the US Patent and Trademark Office’s (USPTO) decision that a robot cannot be an inventor, as well as questions to ponder. The article raised a vast array of questions, including topics such as AI independence and identity, AI ownership of a patent, and AI rights under a legal contract.
Now, the story continues with a recent development. On September 2, 2021, the USPTO’s rejection was upheld by the US District Court for the Eastern District of Virginia. The invention was considered unique, and if Dr. Thaler had applied as the inventor, he would’ve most likely had the patent approved.
But with AI listed as the inventor, the legalities became evident. US District Judge Leonie Brinkema concluded that a robot cannot be an inventor based on definitions from the US Patent and Trademark Office, as well as precedents set by previous court cases.
Brinkema provided the decision in three parts. The first part, conceded by both sides, stated that if DABUS could not be labeled as the inventor, the patent would be rejected. The second part was more controversial. Here, the judge used the definition of an “individual” from the USPTO. The USPTO pulled their definition from Title 35 of the United States Code. “Individual” is defined as a “natural person.”
AI was not considered a natural person and therefore, given today’s technological limits, does not possess a mind wherein its own creative ideas can be formulated. Moreover, she specified that the word “individuals” intended use was to promote equality between male and female inventors.
As further evidence, Brinkema cited a case, Beech Aircraft Corp. v. EDO Corp., in which a patent was denied because an entire corporation was labeled as the inventor. Again, that ruling mirrored DABUS in that it cited an “individual” as a “natural person.” Finally, this brought the judge to the third part of her conclusion. She emphasized that the current legislation is unable to verify DABUS as the inventor. Consequently, in the future, legislators and AI experts must come together to amend the laws to allow AI to become classified as an inventor.
This decision only answered a single question, though. Does the current legislation allow for AI to be considered an inventor?
As of right now, the answer is a simple “no.” But this consideration opens the door to a trickier path. The USPTO requires the inventor’s mind to have come up with the invention. What defines a mind? When can a robot be considered to have a mind of its own? There are many AIs that have human-level intelligence and human-level personalities. Humor, sadness, anger, and happiness can all be programmed into a robot’s “mind,” but when do the robot’s decisions become unprogrammed and solely from learned behavior? Aren’t all decisions that humans make programmed by our past choices, experiences, and lessons?
Dr. Thaler also raised another good point of discussion. His grounds for his steadfast claim that DABUS imagined the invention while conceding that if DABUS was not seen as the inventor, the patent should be rejected, were very basic. They stemmed from his moral belief that it would be deceitful to sign the required USPTO contract affirming the idea was his own.
Thaler had no experience working with food containers or lights. He did not come up with the idea nor did he program AI to come up with that idea. DABUS was only told solve a specific problem using a unique invention. And it did just that. So, if the AI does not have a mind of its own, and the idea was not derived from the mind of Dr. Thaler, who came up with the invention?
This topic of AI’s inventorship classification is relevant in India as well. Look at the current state of India’s patent process. The difficulties and challenges faced by patent applicants have spurred scrutiny of existing legislation. Looking ahead, many speculate that there will be changes throughout the system to make India’s patent practices less complicated.
The legislators of today must be prepared to answer the questions of tomorrow. What is India’s response to robotic inventors? Will they follow suit with the US, UK, and Europe, or will they side with Australia, accepting potential AI-invented patents? COVID-19 has continued to rock India and with AI at the forefront of technological solutions, India must strive to solve these problems before they arise.
With Judge Brinkema’s decision fresh in the spotlight, questions like these are next in queue to be answered. The light-speed progression of technology will, undoubtedly, result in an AI inventor one day. But that day is not today. Alas, the robot invasion must be put on hold when it comes to inventions for now.
(Dr. Raj Dave is an IIT expert-turned patent attorney based in the US and President of Davé Law Group, Virginia, USA. Om Desai is a Technical Specialist at Davé Law Group.)