This article is more than 1 year old

One man's battle to get patent rights for AI inventors in America may be over

Stephen Thaler is fighting the law, and the law is winning

Analysis Stephen Thaler, founder of software biz Imagination Engines, has waged a years-long campaign fighting for machines to be legally recognized as inventors around the world. Now, it looks as though his struggle may have been for nothing.

The battle started when he attempted to file patent applications on behalf of someone – or rather, something – with an odd name given for the original inventor. First name: "DABUS", last name: "Invention generated by artificial intelligence." 

Thaler claims his AI system DABUS, which stands for "Device for the Autonomous Bootstrapping of Unified Sentience," invented an emergency beacon-like device that flashes with unique patterns, and a food container shaped using fractal geometry. After US, UK, and other government officials rejected his patent applications on behalf of DABUS for those objects, with DABUS listed as the inventor, he started suing.

South Africa is the only nation to accept DABUS as an inventor so far.

With the help of a legal team led by Ryan Abbott, a professor of law and health sciences at the University of Surrey in England, Thaler has filed complaints against the US Patent and Trademark Office (USPTO), the UK's Intellectual Property Office, the European Patent Office, the German Patent and Trade Mark Office, and Australia's Commissioner of Patents, seeking to have his AI system's patents granted.

His first line of attack was to obtain IP rights for DABUS in hope of setting some kind of precedent. After a US federal district court in Virginia sided with the USPTO's decision to reject his patent application, stating only "natural persons" or humans can be inventors, Thaler took his case to the Court of Appeals and lost there too.

Now, the only way to keep fighting the rules in the states is if America's highest judicial body, the Supreme Court, hears his case. Which is unlikely.

"We plan to appeal," Abbott told The Register. "Although the Supreme Court takes a very small percentage of cases received."

The Court of Appeals ruling, issued last month, zeroed in on key legal language in US patent law. "This case presents the question of who, or what, can be an inventor. Specifically, we are asked to decide if an artificial intelligence (AI) software system can be listed as the inventor on a patent application," the court documents stated [PDF]. 

"At first, it might seem that resolving this issue would involve an abstract inquiry into the nature of invention or the rights, if any, of AI systems. In fact, however, we do not need to ponder these metaphysical matters. Instead, our task begins – and ends – with consideration of the applicable definition in the relevant statute," the ruling continued.

In other words: there's no need for a debate over the rights of computers. America's key patent law, the Patent Act, is plain: natural persons only. Thaler's case may well struggle to go anywhere at this rate.


The Leahy–Smith America Invents Act [PDF], signed into law in 2011 to update America's patent rules, also states this plainly: "The term 'inventor' means the individual or, if a joint invention, the individuals collectively who invented or discovered the subject matter of the invention."

The legal definition of an individual means a "human being" or a "person", the appeal judges said. "For instance, the [Patent] Act uses personal pronouns – 'himself' and 'herself' – to refer to an 'individual.' It does not also use 'itself,' which it would have done if Congress intended to permit non-human inventors," they added. 

Abbott, however, pointed out there are times that non-humans have been recognized as individuals: for example, he said, corporations. And so, we're told, if businesses can be considered individuals, there's a chance software can be, as well.

"We disagreed with the federal circuit ruling because, among other things, prior Supreme Court cases had held that individuals could be artificial persons like companies, and the opinion results in an un-owned and non-patentable invention, which is not what the Patent Act was meant to do," he told us.

The Patent Act is useful as the US economy can benefit from public, granted patents, thus Abbott argued the patent system should accept inventions created by machines. 

Thaler strongly believes that giving machines rights will encourage innovation and the generation of ideas, and that hampering this progress means the duties of the Patent Act aren't being fulfilled, which is, in his mind, potentially unconstitutional. The Court of Appeals dismissed these claims, and said it only had to look at the letter of the law, if not the spirit, to make its judgment.

"When a statute unambiguously and directly answers the question before us, our analysis does not stray beyond the plain text. Here, Congress has determined that only a natural person can be an inventor, so AI cannot be. Accordingly, the decision of the district court is affirmed," it said.

Thaler told The Register that blocking machines from being inventors means that any IP generated by software will not be made public, thereby locking the benefits of technological progress. "[Imagine] you've just invested a small fortune for an inventive AI system and it invents a cure for cancer. Do you keep it a trade secret or risk losing it to the public domain due to a failed patent application that lacks a named human inventor?" he said hypothetically.

He also believes that if computers can be recognized as inventors, humans will be able to steal their ideas and apply for patents in their own name using those ideas, which will discourage the use of machine learning to come up with patentable technologies.

Moving toward the fringe, however, Thaler also insisted that DABUS was conscious and sentient and had the potential to invent technologies that could tackle "a range of economic, ecological, and sociological challenges."

Before you scoff...

It may be easy to dismiss Thaler, though his efforts to challenge and poke his way through the law are rooted in the possibility that artificial general intelligence could be developed in our lifetime, something experts also take seriously.

At some point in the future machine capabilities may overtake humans across many general tasks, and have a major impact on the way we live and work. In some cases, the future isn't that difficult to imagine. AI-generated art, for example, is already alarming some artists, illustrators, and photographers, who believe computers could start taking away their jobs.

Some even go as far to believe that one of the greatest challenges for humanity is making sure AI is harnessed for good, Thaler argued. Recognizing machines as inventors is one step toward making sure society will benefit from its creations, he said. 

"Machines can outpace humans in generating intellectual property, thus producing unprecedented global prosperity. That could lead the way to an economy in which people strive to improve themselves rather than brutally competing against others for wealth and resources. I believe that would amount to a better, but still not perfect world," he said.

It's not clear when, or if, AI will ever reach the heights Thaler envisions or whether its abilities will forever remain within the realm of science fiction. The courts are not swayed by arguments of AI being on a par with people so far, but even if the US Supreme Court doesn't take on Thaler's case, he won't be giving up.

Thaler is still working with Abbott for an upcoming hearing with the UK's Supreme Court, and still has cases pending in Germany, Australia, and with the EU. If that fails, Thaler is still waiting to see whether his patents for DABUS filed in 11 other countries get approved. If they don't, he might just file yet another lawsuit. ®

More about


Send us news

Other stories you might like