Experts say AI is now present in everything from cell phone apps and facial recognition software to ChatGPT, which allows users to ask questions and receive human-like replies based on data harvested from the internet.
But as usage of AI technology has grown, public concern has been mounting that it could one day surpass human control, lead to widespread unemployment, and worsen racism and racial stereotypes.
As Congress considers bills to regulate the technology, some states have also begun passing laws aimed at placing limits on the technology's usage.
Efforts to regulate AI usage in Oklahoma have largely stalled, and some lawmakers are concerned that the Legislature is now faced with playing catch up.
Experts say previous regulatory efforts have been complicated by the sheer complexity of the issue and by pushback from technology industry lobbyists who oppose government interference in the burgeoning industry.
State lawmakers, meanwhile, have held two different interim studies dealing with AI this month. But the studies highlighted that they're still grappling with the pros and cons of the technology, what to regulate and where to begin.
Legislative action
The Oklahoma Legislature has introduced 11 AI-related bills since 2003, but adopted just one, Cody Allen, of the Southern Council of State Governments, said during a Senate study Monday.
Senate Bill 100 was adopted in 2019 to require eye patients to be assessed and treated by an optometrist rather than by software systems alone, Allen said.
Numerous states have adopted AI-related laws since 2019, and over 200 bills were proposed in 2023, Allen said. He also said some states use AI to improve roads, reduce wait times in health care settings and to develop school safety programs.
Sen. Dave Rader, R-Tulsa, who requested an interim study, said Oklahoma is playing catch up on legislation.
"Where is that line between free market and some type of state involvement?" Rader asked. "You want to make the market free, but you want to protect against the bad guys out there, too."
Gov. Kevin Stitt formed an AI task force in September to examine the technology's abilities and security risks. A report from the task force is due by Dec. 31. Several other states have formed similar groups to study and recommend legislation.
Recent bipartisan AI bills have faced pushback in Oklahoma's Legislature.
Former Rep. Collin Walke and Rep. Josh West, R-Grove introduced data privacy bills in 2021 and 2022. The bills won support in the House, but failed to receive a hearing in the Senate. West introduced a similar measure, House Bill 1030, this year, but it didn't get heard in the Senate.
West said the bill is designed to prevent artificial intelligence from being used for disinformation and cyberattacks. It also would have required companies to request the user's permission before obtaining and using personal information.
He said lobbyists from big tech companies and the Oklahoma State Chamber pushed back against those bills over fears it would overreach into the free market.
"There's no free market that exists without rules," West said.
The State Chamber did not respond to a request for comment.
Regulation hurdles
Walke said companies are scared of losing money. When Apple allowed iPhone users to limit app tracking, Facebook claimed it would lose $10 billion in revenue as a result, he said.
Walke, an attorney, is a certified artificial intelligence auditor. He helps companies ensure they are compliant with the European Union's AI Act, which requires businesses to have an ethics committee, security protocols and conduct bias testing, he said.
"This is something that's happening today, people's lives are being affected today — not five years down the road," Walke said.
Regulating the myriad of industries using the technology is difficult because a law applied to one market won't work for another, Walke said. For example, regulations for facial recognition software won't apply to self-driving cars, he said.
"Good luck getting that many bills passed," Walke said.
He said a broader approach to regulation such as an AI bill of rights is likely a better place to start.
"I'm all for it as long as it's got teeth," he said. "But I'd be interested in seeing if they could pass something that's got teeth."
Definitions in a bill of rights will be important, especially for things like bias in hiring and AI algorithms that prioritize care, assess risk or characterize people based on race. Walke said interpretations vary.
Industry specific laws are also emerging in some states.
Alabama adopted Senate Bill 56 in 2022, which banned facial recognition technology as the sole evidence to support arrest and search warrants.
New York City passed a law that requires employers using AI hiring software to have the tool independently audited for bias.
Without restrictions, Walke said it's hard to know if an AI program is treating people fairly.
© 2023 The Norman Transcript (Norman, Okla.). Distributed by Tribune Content Agency, LLC.
