top stories

Schumer Lays Out Process to Tackle A.I., Without Endorsing Specific Plans


Senator Chuck Schumer of New York, the majority leader, laid out a long-awaited framework on Wednesday to regulate artificial intelligence, hoping to create a path for lawmakers to adopt guardrails many industry insiders say are needed on a technology many members of Congress admit they do not understand.

Declaring that Congress “must join the A.I. revolution,” Mr. Schumer steered clear of endorsing any specific bills, instead calling for an approach to A.I. prioritizing objectives like security, accountability and innovation. Mr. Schumer, who predicted that his plan could produce legislation within months, instead is seeking to give lawmakers a comprehensive crash course in A.I. in a setting where partisan rancor might be set aside, before they try to impose rules on the rapidly changing industry.

“In many ways we’re starting from scratch, but I believe Congress is up to the challenge,” he said during a speech at the Center for Strategic and International Studies, adding that “A.I. moves so quickly and changes at near-exponential speed and there’s such little legislative history on this issue, so a new process is called for.”

That new process centers on what he called “first of their kind” listening sessions in the fall, in which lawmakers could learn about the potential and risks posed by artificial intelligence technology from industry executives, academics, civil rights activists and other stakeholders.

Mr. Schumer’s framework lends new gravity — and potentially new organization — to efforts to regulate A.I. as rapid recent advances have underscored both its extraordinary promise and its potential perils. Lawmakers have already held a flurry of hearings and filed bills to create everything from increased transparency requirements for A.I. platforms to restrictions preventing the technology from being used to deploy nuclear weapons, but have failed to coalesce around any sweeping policy.

Still, the reaction to Mr. Schumer’s proposal was mixed. Some experts worried that the listening sessions, which Mr. Schumer called “insight forums,” might slow down the efforts already underway to regulate A.I.

“On the one hand, he is the most influential legislative figure to show some affirmative interest in this topic, and that is a positive step,” said Ben Winters, senior counsel with the Electronic Privacy Information Center, which advocates for comprehensive safeguards against potential dangers posed by artificial intelligence technology. But overall, Mr. Winters declared Mr. Schumer’s fresh-start approach “frustrating and disappointing,” expressing concern that “other stronger, more protective A.I. laws may get sidelined or delayed as the process plays out.”

Industry leaders have warned that A.I. technology could pose an “existential threat” to humanity. But the proliferation of A.I. has also sparked a series of other concerns about its collection of personal data, spreading of misinformation and perpetuation of discrimination. A.I. technology could also have enormous consequences for the global economy, as increasing automation could potentially eliminate millions of jobs.

In many ways, the A.I. debate in Congress is lagging progress in other government forums. The United States signed on to the Organization for Economic Cooperation and Development’s A.I. principles in 2019; last year, the White House published a series of regulatory proposals in a “Blueprint for an A.I. Bill of Rights.” Just this month, the European Union took an important step toward passing a major law to regulate A.I.

On Capitol Hill, however, many lawmakers agree with Mr. Schumer that to develop a sweeping framework for regulation, Congress must first have a better grasp of the issue.

“It’s critical that if we contemplate regulating A.I., which I think most people agree is going to be necessary, we need to understand why we are regulating it,” said Representative Jay Obernolte, Republican of California and a video game developer with a master’s degree in artificial intelligence. He said Mr. Schumer’s framework would be “helpful, because it’s stimulating a discussion” that is necessary.

Mr. Schumer is not the first congressional leader to try to get Congress up to speed on the rapidly expanding implications of A.I. This spring, Speaker Kevin McCarthy, Republican of California, and Representative Hakeem Jeffries of New York, the minority leader, convened a bipartisan briefing for members to discuss the challenges of A.I. regulation with experts; Mr. Schumer is offering senators a similar series of audiences this month.

But while the approach has been systematically bipartisan, it has not succeeded in bridging key gaps between the parties about how to approach regulation.

Several Democrats, for example, have called for a new federal agency to regulate A.I., much like the Food and Drug Administration regulates the agricultural and medical industries. Senator Michael Bennet, Democrat of Colorado, who advocates such an approach, suggested that such a body could “provide the kind of expertise and oversight” necessary to take on large technology companies, since “Congress is never going to do it on its own.”

But the suggestion is anathema for Republicans like Mr. Obernolte, who rejects the idea of potentially duplicating the efforts of federal agencies “that are already grappling with the problem of how to establish rules regarding A.I. within their sectoral spaces.”

There is a similar, unresolved debate in Congress about whether A.I. laws ought to be comprehensive, or pinpointed to address certain topics.

Mr. Schumer insisted Wednesday that his framework is not meant to supersede or hinder efforts to build bipartisan consensus around A.I. legislation in other forums, painting his framework as complementary to the traditional committee process of drafting bills, which he said will “play a central role, but won’t on their own suffice.”

He also stressed that it would be imperative to exercise “humility” when measuring the success of his efforts.

“We’re going to work very hard to come up with comprehensive legislation, because this is so important; we’re going to do everything we can to succeed, but success is not guaranteed,” he said. “It may be exceedingly difficult for legislation to tackle every single issue.”



Source link