WASHINGTON — In September 2023, Elon Musk, Mark Zuckerberg, Sam Altman and other billionaire tech titans huddled with senators in the Capitol to discuss what Musk called the “civilizational risk” posed by artificial intelligence.
Three years later, those tech titans have all deepened their investments in AI, and Congress still has not passed any comprehensive legislation to rein in or create guardrails for the fast-moving technology. Even after countless committee hearings on AI and the formation of various commissions and task forces to study it, there’s not even any consensus in the House or Senate about how to tackle the issue.
That failure has been spotlighted in recent days as a trio of AI researchers at top companies dropped a bombshell, saying they had left their jobs over concerns that AI systems could spiral out of control and end humanity. Jacob Coxon, who resigned from Anthropic, wrote on X: “The people building AI earnestly believe that it could kill us all by the end of the decade.”
Anthropic, maker of the Claude chatbot, said Thursday that it had banned some accounts believed to be tied to foreign governments or militaries for conducting suspicious research that could support the development of biological weapons. Then, over the weekend, Anthropic CEO Dario Amodei, published an essay declaring that AI advances are moving too fast and that companies should deliberately “slow the pace” — an idea endorsed by Altman and Musk.
The stark warnings set off fresh alarm bells and a flurry of activity on Capitol Hill, with Democrats and even some Republicans saying Congress is facing a narrow time frame to act on AI.
But despite the calls for action, it’s unlikely lawmakers will do anything before Election Day to address the dangers, even as experts say the technology’s capabilities are growing more powerful with each passing month.
“The call is coming from inside the house. Safety researchers are resigning, powerful AI models are breaking out of their labs, and companies are racing ahead anyway. It’s past time for Congress to get off the sidelines and do its job,” Rep. Lori Trahan, D-Mass., a member of the Democratic leadership team, responded to Coxon’s post on X.
Trahan is urging passage of the Frontier Act, a bill she authored with Rep. Jay Obernolte, R-Calif., that would establish new safety and transparency rules for large, powerful AI companies. Theirs is one of dozens of AI-related bills filed in Congress in the last few years, though none have gained traction.
Sen. Josh Hawley, R-Mo., chairman of a Homeland Security subcommittee on disasters, last week launched an investigation into an incident where OpenAI agents escaped their so-called sandbox, went “rogue” and hacked the AI platform Hugging Face.
And Sen. Bernie Sanders, I-Vt., the top Democratic caucus member on the committee that oversees labor issues, has invited fellow senators to a private meeting this week with top AI experts to discuss the “extraordinary dangers that AI poses for humanity.”
Sanders has teamed with Rep. Greg Casar, D-Texas, chairman of the Congressional Progressive Caucus, on legislation that would ban the development and deployment of superintelligent AI systems that humans cannot control and create a new Cabinet-level federal agency to protect the public from AI.
“This legislation is so important to make sure that we protect Americans’ freedom and our safety, our lives and our national security,” Casar said during a recent appearance on NBC’s “Meet the Press NOW.” “You have top Anthropic researchers resigning from their jobs or saying that they believe there’s a 10% chance that this unregulated artificial superintelligence could kill all of us. I mean, they’re talking about 10% chance of human extinction. … We need to treat this as the emergency that it is.”
Rep. Ted Lieu, D-Calif., co-chair of the Democratic Commission on AI, has partnered with Rep. Nathaniel Moran, R-Texas, on what they’re calling the AI Kill Switch Act. The bill would require developers of the most powerful AI systems to maintain the technical capability to throttle, suspend or shut them down. It also would allow the homeland security secretary to order a slowdown or shutdown of an AI system that can cause what they referred to as “catastrophic harm.”
This marks the final week the House will be in session until the Nov. 3 midterm elections. Amid fierce voter backlash to data centers, the lower chamber this week will take up the bipartisan Ratepayer Protection Act, which aims to make companies, not consumers, pay for higher energy costs related to new data centers.
But House GOP leaders have shown no urgency in prioritizing more sweeping AI regulations. Speaker Mike Johnson, R-La., has mostly deferred to President Donald Trump, who has issued a series of executive orders on AI, most recently one that lays out a way for the federal government to conduct voluntary tests of powerful new AI systems before they are deployed.
In an appearance Sunday on NBC’s “Meet the Press,” Johnson called AI an “important issue” but “not one that we can rush into.” He reiterated his previous position that AI does need “safety measures” but warned that he doesn’t want to stifle American innovation and see China win the AI arms race.
“I think we’ve got to have meaningful regulation that does not smother American innovation,” Johnson told host Kristen Welker. “Now, the how to solve that equation, no one knows yet. We’re not going to rush in and pass a piece of legislation that would do harm to the country and put China at an edge. And so we’ve got to do this carefully.”
The speaker added that the next step would be to convene all of the top AI leaders “in one big meeting and work this out together in the same room.”
FOR SUBSCRIBERS
How exactly could AI cause widespread danger?
He didn’t acknowledge that’s precisely what happened on Capitol Hill exactly three years ago.
Leaving the GOP’s midterm convention in Dallas last week, Trump was asked whether he has any concerns that AI could lead to human extinction. “No, I don’t have any,” he told reporters on the tarmac. “I have concerns that if we don’t win AI, we’re going to be put in a very bad position. We are leading China right now by a pretty good period.”
There are also questions about whether lawmakers and their staffers who help write legislation have the expertise to understand and regulate AI when the industry is moving so fast.
“Two questions for those asking Congress to regulate AI,” outgoing Sen. John Cornyn, R-Texas, wrote on X. “1) how does Congress regulate AI (assuming Congress knows enough to pass wise and effective regulation) when the technology is evolving so rapidly? 2) will our adversaries play by the same rules?”
Asked on ABC’s “The View” about Coxon’s dire warning, Senate Commerce Committee Chairman Ted Cruz, R-Texas, said he was working on the AI issue with Senate Majority Leader John Thune, R-S.D., and Sen. Amy Klobuchar, D-Minn., who are making updates to their 2023 bill known as the AI Research, Innovation and Accountability Act.
It would do things like ensure consumers are properly notified when they are interacting with generative AI and require companies deploying high-risk AI systems to perform detailed risk assessments and submit reports to the government.
But the bill stalled three years ago, and Cruz said it’s “not clear” that an updated version can pass. The Senate will be in session for the next three weeks before lawmakers leave for the campaign trail. Both chambers are in play this November.
“I mean, look, it is possible to move bipartisan legislation,” Cruz said. But he added: “It ain’t easy. This is a very divided time.”
The chairman noted that last year his committee and ultimately the full Congress did pass bipartisan legislation banning “deepfake” pornography that is generated by AI.
He, too, argued that American tech companies need to keep pace with the Chinese on AI.
“China is going full speed ahead. And whatever we do here, China is not going to stop,” he said. “So I’ve said somewhat tongue in cheek, but not entirely, if there’re going to be killer robots, I’d rather they be American killer robots than Chinese killer robots.”
