In the ongoing battle between robots and human inventors, score one for natural-born people: Last month, a federal appellate court held that artificial intelligence (AI) cannot, under American patent law, be regarded as an inventor.

The long-anticipated decision wasn’t much of a surprise, though it came as a disappointment to Stephen Thaler, the polymathic inventor who argued that his own machine, “DABUS,” the Device for the Autonomous Bootstrapping of Unified Sentience, deserved credit for inventing numerous technologies; Thaler maintains that he himself made no inventive contribution to DABUS’s own independent innovations.

As discussed in this space on previous occasions, Thaler filed patent applications in Europe, Israel, South Africa, Taiwan, the UK, the US, Vietnam, and other locations, claiming inventions that spanned an emergency warning light to a food container graspable by robots. These attempts, however, have generally been met with resistance from patent offices across the globe, which have relied on patent statutes requiring an actual human to be named as inventor. (South Africa appears to be the only country thus far that has granted patents to DABUS; Israel’s decision is still pending.)

So, too, was DABUS’s patent application initially rejected by the United States Patent and Trademark Office (USPTO), which concluded that the applicable US patent legislation limits the definition of “inventor” to natural persons. Thaler challenged the ruling in federal district court in Virginia, which sided with the USPTO, and then appealed to the US Court of Appeals for the Federal Circuit, which has exclusive jurisdiction over patent-related appeals.

And so in August, the Federal Circuit held that “there is no ambiguity: the Patent Act requires that inventors must be natural persons; that is, human beings.” Specifically, Judge Leonard Stark, writing for a unanimous three-judge panel of the court, ruled that the statute repeatedly refers to inventors as “individuals,” and that the US Supreme Court and commonly used dictionaries define “individuals” as “human beings.”

The Federal Circuit also held that the patent statute requires an inventor to submit a declaration that “such individual believes himself or herself to be the original inventor,” and that “while we do not decide whether an AI system can form beliefs, nothing in our record shows that one can, as reflected in the fact that Thaler submitted the requisite statements himself, purportedly on DABUS’ behalf.”

Likewise, the court rejected Thaler’s argument that the patent statute elsewhere uses the broader term “whoever,” noting that those contexts relate not to who made an invention but to how it was made or to who may be infringing it. Just as corporations cannot be credited as inventors, the Federal Circuit concluded, so too is AI barred from inventorship.

So what’s next? Thaler plans to petition for review from the Supreme Court, according to his attorney, Ryan Abbott, who told Bloomberg Law that the Federal Circuit decision “ignores the purpose of the Patent Act and the outcome that AI-generated inventions are now unpatentable in the United States. That is an outcome with real negative social consequences.”

Indeed, as I’ve argued here and elsewhere, from a policy perspective, recognizing AI inventions would afford significant potential benefits to our patent system and would help spur innovation across a variety of sectors. But given the language of the patent statute and the current constellation of the judiciary, new legislation or at least USPTO guidelines may be required to make such recognition a reality. Either way, this isn’t the last we’ll be hearing from DABUS.

The post US Court of Appeals Short-Circuits Artificial Intelligence Inventions appeared first on American Enterprise Institute – AEI.