A newly released report commissioned by the State Department said artificial intelligence is creating entirely new categories of “weapons of mass destruction-like” risks. Those risks, according to the report, include catastrophic events that could lead to human extinction.

The report also warns of the potential consequences if the U.S. government fails to regulate AI. 

“The prospect of inadequate security at frontier AI labs raises the risk that the world’s most advanced AI systems could be stolen from their U.S. developers and then weaponized against U.S. interests,” the report said.   

According to Time, which first published the report, the authors make policy proposals to prevent these negative outcomes. The policy proposals include making it illegal to train AI models using more than a certain level of computing power, making it illegal to publish the inner workings of AI models under open source licenses, and tightening controls on the manufacture and export of AI chips. 

Multiple members of Congress said those ideas sound viable. However, there are major hurdles to making that happen. 

First, as the report states, “development in AI is now so rapid that an ordinary policymaking process could be overtaken by events by the time the resulting policies take effect.” 

Lawmakers admit that they’re learning about AI as they go, so there is a curve to overcome before they can write effective legislation. 

The committees of Congress that are given the responsibility of viewing AI are hardly up to the challenge and the task in terms of their own knowledge and information available to them.

Sen. Dick Durbin

“I’m not an expert in the area and that’s one of the problems,” Sen. Dick Durbin, D-Ill., said. “The committees of Congress that are given the responsibility of viewing AI are hardly up to the challenge and the task in terms of their own knowledge and information available to them. You know, that’s what has bothered me the most when you give this assignment to Congress, you have to be prepared for a long lead in period.”

There is a Senate artificial intelligence working group that has been tasked with taking lead on the issue. The group is trying to pass major legislation this year. 

The second hurdle is Big Tech opposition. Sen. Josh Hawley, R-Mo., said Big Tech has too much influence in Congress and it will try to squash legislation that isn’t in their favor. 

“Do you think any AI legislation is going to see any time of day on the Senate floor,” Hawley questioned. “Not in this Senate. Because why? Who owns the AI technology and is developing it? The same mega corporations, Google, Microsoft, Meta, you know, TikTok, probably, too. I mean, these guys, if they don’t want it, it doesn’t see time on the Senate floor. It doesn’t matter who’s in charge. Democrats, Republicans, doesn’t matter. All of them bought and paid for by these corporations.”

Hawley, along with Sen. Richard Blumenthal, D-Conn., have a bipartisan framework for AI legislation which would establish a licensing regime administered by an independent oversight body and ensure legal accountability for harms.

Blumenthal pointed out the third hurdle: the U.S. is already behind in its AI capabilities. He said prime examples are the chips and semiconductors in Russian weapons used against Ukraine. 

“The Ukrainians are finding Russian chips, AI components, that are now a fact of modern warfare,” Blumenthal said. “[AI] is exploding exponentially in its defense uses and ought to deeply concern, in fact, gravely frighten the American people because we are falling behind in AI technology, particularly in defense uses and it’s a matter of national security.”

There are more than 75 AI-related bill proposals in Congress right now.