Will AI Take Over the Process of Invention?

Artificial intelligence (AI) development has made leaps and bounds in recent years — to the point that the technology is now assisting inventors in a variety of ways, such as identifying promising new materials or drugs. But what happens when AI is solely responsible for the actual act of invention itself?

An Artificial Inventor

Ryan Abbott is a patent attorney and health sciences professor at the UK’s University of Surrey, located about an hour away from London. And he wants to get to the bottom of this AI-invention dilemma. “If I write a Word document with Microsoft Word, that doesn’t make Microsoft Word an author, and if use an Excel spreadsheet, that’s not making Excel an inventor of a patent I make,” he explains.

Abbott is also one of the lawyers working on the Artificial Inventor Project (AIP), an endeavor that’s seeking “intellectual property rights for the autonomous output of artificial intelligence.” He strongly believes that there are times when it makes the most sense to consider an algorithm or software the actual inventor of something.

Case in point: Last August, AIP filed patents for a food container and a warning light on behalf of Stephen Thaler, CEO of Imagination Engines. But instead of listing Thaler as the author of the applications, Dabus AI, an AI system that Thaler spent more than ten years developing, took that credit. Why? Because Dabus AI concocted these innovations after being fed general information about various topics.

Thaler built Dabus. But he has no expertise in food containers or lights. So he wouldn’t have been able to generate these concepts by himself. As a result, AIP argues that Dabus is the rightful inventor.

While the UK and European patent offices considered both inventions worthy of patents, they recently rejected the applications since the inventor wasn’t a human. Consequently, the devices are not under patent protection.

Unsurprisingly, Abbott is already planning to appeal these decisions. He thinks that it’s inevitable for more cases to arise where AI should be considered the inventor, and the law needs to be prepared to handle them properly. If it isn’t, Abbot believes that unintended consequences could occur which jeopardize the future of innovation.

The Argument for AI Invention

Nobody is arguing that AI should own these patents. It’s quite common for an individual to be listed as a patent’s inventor while its actual owner is the company employing the inventor. So, in this case, AIP is claiming that Dabus AI is the inventor, and Thaler is the owner.

Simple enough, right? Well, it gets more convoluted when you consider that patent law assigns ownership in particular ways. For instance, the listed inventor must either be an employee or contractor of the owning party. And according to Peter Finnie, IP expert at Potter Clarkson, AI fits neither of these legal categories. To make matters worse, inventors are required to be individuals and “natural persons” — another requirement that AI doesn’t meet.

Being an inventor also comes with a unique set of responsibilities, such as being able to enter into contracts, authorize licenses, and file lawsuits. AI can do none of these things.

“I won’t dispute that AIs are really good at solving problems and solving them in ways that are new and different and that people could maybe never come up with,” says Chris Mammen, IP lawyer at Womble Bond Dickinson. “But as a policy matter, I’m not sure that our patent system is the right tool to reward the development of those kinds of solutions.”

Yes, AI and machine learning applications are now making major contributions to innovation. It’s not even that farfetched to imagine a scenario in which humans make play no role in the innovation process. But we’ve still got a long way to go before the law aligns with this new reality. That’s not to say it won’t happen some day.

A New Era of Invention Is Inevitable

While Abbott stresses that society must figure out this conundrum sooner than later, he does acknowledge that AI doesn’t just come out of nowhere. This technology must be coded, trained, and fed data. But does this mean that everything an AI creates can and should be traced back to humans? Such a feat quickly gets convoluted.

Hundreds, if not thousands, of people could be involved in the construction of an AI project such as IBM’s supercomputer Watson. But “if Watson then applies those capabilities and solves a particular problem in a way that results in a patent, it’s not clear that anything any of those people have done qualifies them to be an inventor,” Abbott explains.

So in cases like Thaler’s and Dabus AI’s, if humans aren’t intimately involved in the invention process, and AI can’t be listed as the inventor, the invention may not be able to be patented at all. Obviously, this is problematic and could cause organizations to stop investing money in AI. As a result, potential breakthroughs in life-changing areas like drug discovery may never happen.

A spokesperson for the European Patent Office says AI will most likely be considered a tool instead of an inventor for the foreseeable future. Jeremy Smith, IP attorney at Mathys & Squire, says that it’s possible that the law will change to accommodate AI. Maybe the tech won’t be listed as an inventor, but perhaps a new patent-like right could take its contribution into account.

Smith is quick to point out that lawmakers move extremely slowly. Abbott agrees and knows this situation will most likely take months, if not years, to figure out. But one thing’s clear: AI is disrupting the invention process in such a way that it will never be the same again.