It’s unlikely there are many members of Congress as knowledgeable about the implications of artificial intelligence as Rep. Jay Obernolte (R-CA).
Obernolte is the chair of the bipartisan AI Task Force, which is working on a major report on AI to be released by the end of the year.
“It’s pretty clear artificial intelligence is going to change the world,” Obernolte said during WTOP’s Road to Responsible AI.
He has spent considerable time thinking about AI, ever since high school when his ambition was to be a researcher in AI. Obernolte acknowledged that it’s “very strange symmetry” that he’s now the congressional point person on AI, given his background.
He completed a degree in computer engineering at the California Institute of Technology and was working on his doctorate in AI at University of California Los Angeles. But then his side hustle writing video game software took off, and he ventured away from academia and into business.
“Ultimately, I got into politics and found myself in Congress,” Obernolte said.
He’s proud of the work the AI Task Force has done so far, including more than 15 hearings on a variety of AI topics.
The Democratic co-chair of the task force is a fellow Californian, Rep. Ted Lieu.
“I am cautiously optimistic that our report will be the furthest thinking document on AI that’s been produced by the legislative branch so far,” Obernolte said.
Public misconceptions about AI
As someone who has been immersed in AI for much of his life, Obernolte is familiar with the concerns that have been raised about its use — and potential misuse.
He said many impressions of AI have been “informed or misinformed” by decades of science fiction and pop culture.
“If you ask the average American, what are the chief dangers of AI? What is the maximum potential downside? You’ll get an answer out of a ‘Terminator’ movie where an army of evil robots rises up to take over the world with … their red laser eyes,” he said.
But he said the task force and lawmakers look at a lot of serious issues: the potential spread of disinformation, national security implications, potential malicious use of AI to pierce digital data privacy and use of AI to conduct cyber fraud.
“We call it the parade of horribles and suss out what is government’s responsibility to guard against the public, guard the public against those threats and then to recommend a path for carrying out that action,” he said.
The task force has 24 members “who are highly engaged and educated” on the issue, Obernolte said, adding that’s important as the task force moves forward on trying to coordinate AI regulations with federal agencies.
Don’t ‘regulate for the sake of regulating’
Obernolte, a Republican, and task force member Virginia Rep. Don Beyer, a Democrat, agree that regulation of AI needs to avoid what they view as an onerous bureaucratic approach that’s been adopted by the European Union.
Obernolte said the congressional task force is recommending a different path. “We call it the ‘hub and spoke’ approach,” he said. “We believe in sectoral regulation.”
He noted that many in the United States believe that AI is currently unregulated and that the country is well behind other foreign governments.
Obernolte pointed out that the Food and Drug Administration has already processed more than a thousand applications for the use of AI and medical devices.
“You can’t come up with a risk context for usage more consequential than a medical device that’s being used to make clinical decisions for health care or being implanted in someone’s body, and yet the FDA is already having to navigate this space,” he said,
He also pointed out that too much federal regulation can undermine developments in the private sector. While large companies like Apple and Google have plenty of lawyers to navigate legal matters, small companies don’t have those resources.
Over-regulation can hurt “two guys trying to start the next Google in a garage somewhere,” he said.
“I think it’s important to point out to people that you can’t just regulate for the sake of regulating,” he added.
How can deep fakes in political ads be regulated?
Obernolte said there is a need to prevent the misuse of deep fakes — making something look and appear human when it’s not.
Concerns have been raised about fake social media posts that can make it appear a politician has said something they never actually said.
“We could all agree as a society that that’s a bad thing and that shouldn’t be allowed,” he said. “But it’s where the rubber hits the road that matters, which is, how do you articulate that in legislation?”
He acknowledged it’s complicated.
“For example, if a political opponent were to use generative AI to create an image of Darth Vader with his arm around me and publicize that … they would argue that that is parody, right?” he said.
He also noted it’s free speech. Would a reasonable person believe that Darth Vader was really in that picture with him? “I think that’s what gets at the crux of the argument,” he said.
Although it’s a complicated issue, he believes the task force is well on its way to getting its arms around the problem.
Task force report coming soon
Obernolte is looking forward to the release of the task force’s report.
He said the aforementioned issue touches on matters involving intellectual property.
“When our report comes out … one of the most voluminous sections is the one on intellectual property because there are a lot of very thorny intellectual property issues that we’re going to have to solve with respect to AI, and this is just one of them,” he said.
In addition to the report work, the task force has already backed numerous bills related to AI.
He said that reflects one of the “pillars of the report,” which is to embrace the concept of incrementalism.
“AI is such a technologically diverse field. It is so fast-moving, it is absolutely impossible to pass one piece of regulation today, this year, and call it good,” Obernolte said.
Among the bills that he hopes to see passed is the CREATE AI Act, which would establish resources to help small business working to develop cutting-edge AI systems that are safe and transparent.
“It’s also important to give the American public some confidence that Congress is capable of dealing with this issue,” he said. “And if we can get some bills actually signed into law that make those meaningful first steps, I think we will demonstrate that.”
Discover more articles and videos now on WTOP’s Road to Responsible AI event page.
