Congress wants to regulate AI, but it has a lot of catching up to do

For the past several weeks, Senate Majority Leader Chuck Schumer has met with at least 100 experts in artificial intelligence to craft groundbreaking legislation to install safeguards.

The New York Democrat is in the earliest stages of talking to members of his own party and Republicans to gauge their interest in getting behind a new proposed AI law.

"Our goal is to maximize the good that can come of [artificial intelligence]," Schumer said. "And there can be tremendous good, but minimize the bad that can come of it. ... But to do it is more easier said than done."

It's all part of a congressional race to try to catch up legislatively to exploding advances in AI.

Monday night, a bipartisan group of House members will host a top industry figure for a joint dinner. On Tuesday, a Senate panel will hold a hearing to consider new oversight of the technology.

But while lawmakers look to craft AI rules, they face Congress' lackluster history of regulating emerging technologies.

Schumer admits he's facing some clear challenges.

"It's a very difficult issue, AI, because a) it's moving so quickly and b) because it's so vast and changing so quickly," Schumer said.

As Schumer works to build a bipartisan consensus behind his legislative framework, he must also navigate a bitterly divided Congress.

Congress has struggled to regulate emerging technology

Congressional lawmakers missed critical windows to install guardrails for the internet and social media.

Now, it faces the equivalent of trying to put in brakes for a runaway train.

"AI, or automated decision-making technologies, are advancing at breakneck speed," said law professor Ifeoma Ajunwa, who co-founded an AI research program at the University of North Carolina at Chapel Hill. "There is this AI race ... yet ... the regulations are not keeping pace."

Ajunwa says that there aren't enough experts in both computer science and law on Capitol Hill and that this makes AI lawmaking all the more challenging.

"There is a real need for such ... legal training for people who will then end up in Congress," she said.

The name of the technology alone can add a mystique and cause confusion for lawmakers, Ajunwa argues.

She and some industry experts say AI should instead be called "automated decision-making" to reflect the human decision-making — including values and biases — embedded in it.

Lawmakers play catch-up

Sen. Josh Hawley, R-Mo., wants to play a role in the development of AI law, but he admits he has work to do.

"I've got to get educated," he said during a recent ride on a Senate subway train back to his office.

Hawley has loomed large in partisan fights over a variety of issues, but on this topic, he's intrigued by the Democratic leader's plans.

Hawley is the top Republican on a Senate Judiciary Committee subpanel that will examine AI oversight options in a hearing on Tuesday.

"For me right now, the power of AI to influence elections is a huge concern," Hawley said. "So I think we've got to figure out what is the threat level there, and then what can we reasonably do about it?"

Sen. Richard Blumenthal, D-Conn., chairs the subpanel that will hold Tuesday's hearing. Sam Altman, CEO of OpenAI, the company behind the chatbot ChatGPT, will testify for the first time before a congressional panel.

But that marks just one of many planned AI hearings.

Sen. Gary Peters, D-Mich., who chairs the Senate Homeland Security and Governmental Affairs Committee, plans to hold at least one hearing on AI during every work period.

Peters argues that Congress has already seen some progress passing legislation related to AI, including four bills that Peters wrote during the last Congress.

"We're going to continue to focus on that in Homeland Security," he said. "We had a hearing last month. We're going to have another one coming up later this month."

Across the Capitol, Rep. Ted Lieu, D-Calif., will co-lead a bipartisan dinner hosting OpenAI CEO Altman on Monday.

This year, Lieu introduced the first piece of federal legislation written by AI. Lieu said he used ChatGPT by asking the bot how he should write a resolution pushing for AI regulation.

"You have all sorts of harms in the future we don't know about, and so I think Congress should step up and look at ways to regulate," Lieu told reporters just before introducing the legislation.

The urgency is obvious

Law professor Ajunwa, who recently wrote a book on the influence of tech and AI on the modern workplace called The Quantified Worker, worries about AI's privacy issues. She says key questions are not being asked about the technology's impact on disadvantaged people, while the focus remains on profits.

"The way the internet developed, unfortunately, is the same way that AI is developing," she said.

Ajunwa says that with the U.S. already lagging behind the technology — for example, the European Union is already years ahead in regulation efforts — the best bet for regulation may be quicker White House executive actions.

The Biden White House announced a series of initiatives ahead of meetings with industry officials this month.

Still, back at the majority leader's office just off the Senate chamber, Schumer remains undeterred.

"Look, it's probably the most important issue facing our country, our families and humanity in the next 100 years," he said. "And how we deal with AI is going to determine the quality of life for this generation and future generations probably more than anything else."

Copyright 2023 NPR. To see more, visit https://www.npr.org.