WASHINGTON (CN) — Holding up an oddly shaped blue hunk of plastic, University of Surrey law school professor Ryan Abbott told a group of senators that it was indicative of the challenges facing inventors trying to leverage developing artificial intelligence technology.
The blue receptacle, Abbott explained, was a new invention — a 3D-printed mockup of a beverage container based on fractal geometry. Using complex geometric patterns maximizes surface area and makes the vessel easy to grip, he said. It also makes it difficult to drink out of and nearly impossible to clean.
The most unique thing about the fractal-inspired container, though, was that it wasn’t dreamed up by a human inventor, but instead by Abbott’s AI model, known as Dabus.
“Dabus was built and trained by people, but on general knowledge, rather than on knowledge specifically about container design,” Abbott told members of the Senate Judiciary Committee during a hearing Wednesday. “It was not asked to solve a specific problem, and its user had no training in container design or the specific intent to design a new container.”
The idea to create a beverage container originated entirely from within the AI’s neural network, Abbott said, a fact that would make Dabus a patent inventor if it were a human. However, he said, U.S. intellectual property officials would not grant a patent for the AI-made container under current law.
“Troublingly, the United States has emerged as an international outlier in terms of refusing to provide a patent for this invention,” Abbott said, “not because the invention itself isn’t deserving of protections, but because of how it was invented.”
Intellectual property issues posed by the fast-developing artificial intelligence industry have been of particular interest to lawmakers in recent months, as Congress aims to get out ahead of the technology by establishing new regulations.
While previous discussions have centered around copyright law and the rights of human creators whose work is parsed by AI models, a panel of experts at Wednesday's hearing focused on legal roadblocks preventing human inventors from using AI to their advantage. One such barrier is a provision unique to U.S. copyright law known as conception law, which requires that an invention be a new idea from the mind of an inventor.
“We must ensure that the doctrine of inventorship law known as conception … is not applied rigidly to deprive human inventors of patent rights, simply because generative AI now allows some of the inventive process to occur on a computer,” said Corey Salsberg, vice president of global intellectual property affairs at pharmaceutical company Novartis.
Conception law, the witnesses explained, poses a challenge to American inventors at a crucial moment if Washington hopes to be a leader in AI technology.
“AI has extraordinary promise in the mid-21st century," said John Villasenor, a professor of electrical engineering and law at the University of California, Los Angeles. “Maintaining U.S. global economic competitiveness will require that the United States be a leader in AI innovation, and a patent system that incentivizes the use of AI to innovate is a key component of that leadership.”
To protect AI-assisted innovation, Congress should amend the longstanding Patent Act to remove language that prohibits patents based on how an invention is made, Abbott suggested. “Designation of inventorship is less important than patentability.”
However, the law should also force would-be patent seekers to disclose that their inventions were made using artificial intelligence, Abbott added. Such a requirement will keep the public informed and prevent people from claiming credit for work created by AI.
Laura Sheridan, head of patent policy at Google, told lawmakers that the U.S. Patent and Trademark Office should also take steps to shore up inventorship rights for AI-assisted technologies, such as technical training for patent review staff.
Google also supports hiking patent fees for large companies while keeping them low for smaller patent applicants, Sheridan said. “We believe that the USPTO will be in the best position to issue robust and reliable patent rights, including in complex emerging technologies like AI, when the fees before the grant of the patent more closely match the costs.”
The witnesses urged lawmakers not to sit on their hands when it came to protecting AI-driven innovation.
“It is critical for Congress to act on this now, not just because of the state of AI today, but because investments need to be made now in the AI and innovation of tomorrow,” Abbott said. “There is a universal consensus that AI is only going to get better at doing this sort of thing, and if the right frameworks are in place, the outcome of that should be significant benefits for everyone.”
Wednesday’s hearing was the second time the Senate Judiciary Committee heard from AI experts. The panel in May grilled Sam Altman, founder of OpenAI, on the need to congressional intervention on growing generative AI models such as his company’s ChatGPT.
On the House side, lawmakers heard from copyright experts and content creators last month — some of whom who said AI models would stifle human creativity.Follow @@BenjaminSWeiss
Read the Top 8
Sign up for the Top 8, a roundup of the day's top stories delivered directly to your inbox Monday through Friday.