The face of the artificial intelligence fever sweeping the US — OpenAI Chief Executive Officer Sam Altman — gripped Washington this spring when he pleaded under oath for the first time for Congress to regulate the technology.
Yet behind the scenes on Capitol Hill, an unexpected crew knew that plea was coming and started to assemble.
Ahead of Altman’s May 16 appearance, Senate Majority Leader Chuck Schumer (D-N.Y.) received count-me-ins from Sens. Todd Young (R-Ind.), Mike Rounds (R-S.D.), and Martin Heinrich (D-N.M.) to help steer the congressional response to AI. They huddled together in the majority leader’s office for the first time a day after Altman testified.
The bipartisan group of four since then is meeting almost weekly over coffee in Schumer’s office or on the phone, racing to formulate rules that protect Americans from AI’s worst threats, such as biological weapons and mass unemployment, but also promote the technology’s potential to cure diseases and curb hunger.
Adding to that challenge: Such legislation would need to pass a deeply divided Congress facing a packed agenda, a largely unsuccessful history of regulating technology, and an upcoming election heightening partisan politics.
Still, in interviews with the senators, former staffers, and close observers, optimism is running high that the rare but mighty alliance can pull off the extraordinarily complex task of confronting AI. With Schumer at the helm so far, they organized three briefings over the summer to educate senators on the technology and are preparing nine so-called insight forums to continue that work for the fall. Major tech and industry leaders, including Meta Platforms Inc.’s Mark Zuckerberg, Tesla Inc.’s Elon Musk and Bill Gates are slated to attend the first gathering on Wednesday.
Bill Gates, CEOs of IBM and Palantir to Join Schumer’s AI Forum
The squad’s efforts pave the way in the coming months to craft legislation aimed at strengthening the US’s competitive edge while keeping the country safe. Each senator brings legislative prowess to the table: Schumer is coming off a series of bipartisan deals, and Young was the chief GOP negotiator he leaned on to shepherd one of them, the CHIPS and Science Act (Public Law 117-167). A veteran, Young is also worried about the threat of AI-powered bioweapons. Rounds’ national security expertise could help address some of AI’s deadliest risks and he’s personally seen AI’s promise to deliver solutions through his late wife’s cancer treatments. Heinrich, an engineer plugged into science and tech issues, founded the Senate’s AI caucus four years ago and championed several early AI measures at the time.
It’s a “very practical, very pragmatic, very bipartisan, compromise-oriented, get-things-done kind of a herd that Schumer has corralled here,” IBM Corp. Policy Lab’s co-director Ryan Hagemann said.
The endeavor will test whether lawmakers can achieve solutions in an era of shrinking compromise—a test they cannot afford to fail, observers warn, as AI’s dangers are too significant to be left ungoverned.
“This recent decision to start this group of four—it’s the way Congress should work,” former Sen. Rob Portman (R-Ohio), who co-founded the AI caucus with Heinrich, said. “We’ve gotta deploy AI in a way that keeps us competitive and safe. We don’t have time for the partisanship.”
“I’m encouraged by it,” Portman added, “But I can’t say that it’s normal.”
‘It’s the Future’
The November debut of OpenAI’s ChatGPT shown a light on the abrupt advancements in generative AI—large language models that can generate text, visuals, and audio—and triggered widespread concerns about the technology’s capabilities. By April—lightning speed by Congress’ standards—Schumer pledged to develop bipartisan AI legislation.
“It’s the future,” Schumer said. “There’s a tremendous opportunity to maximize its potential and a tremendous risk. If you do nothing, it could do damage.”
AI Rules Must Balance Innovation, Safeguards, Schumer Says
In June, Schumer announced a legislative framework. Over the summer, the group hosted three closed Senate briefings featuring academics, defense, and government officials to get members up to speed. The purpose is to make senators comfortable with AI so “it’s not a boogeyman. It’s not like a human being with intelligence,” Rounds said. “It is a machine and it’s a machine that is programmed, and it’s not something that we should be afraid of.”
Lawmakers of both parties and chambers are unveiling proposals that grapple with AI, but Schumer is pursuing an all-hands-on-deck approach, expecting each committee to provide input on how AI impacts their jurisdictions to help craft legislation. In July, he predicted a package would come within “months.”
The senators say they’re feeling confident to move legislation, but acknowledge that the rapidly changing tech might require measured, incremental responses rather than a sweeping, conclusive one.
“It’s a massive task, but we don’t have to solve all challenges or determine every threat or opportunity now,” Young said. “This ought to be an ongoing effort.”
The Indiana Republican said their current strategy is to identify existing laws that can be applied to AI and to adapt the legal framework as needed to ensure guardrails are in place. To do so, they’re considering a wide range of perspectives as AI permeates daily life. The senators also said they want federal agencies to become better versed in AI to address threats and opportunities.
AI Influencers Pound Hill Hallways to Shape Legislation
Another key focus of the group is investing in research, training, and resources toward AI development to maintain US leadership.
“We’ll never be perfect, but every single year we’ll get better,” Rounds said. “Our goal is to make sure that AI development occurs in the United States, and that we do what we can to make sure that our near-peer adversaries have a more difficult time developing AI.”
Tom Williams/CQ-Roll Call, Inc via Getty Images
Sens. Chuck Schumer (D-N.Y.) and Mike Rounds (R-S.D.) during a Senate vote in 2022.
Schumer and Young, Take 2
With AI regulation, Schumer is seeking to replicate last year’s success spearheading a bill that boosted US semiconductor manufacturing. The CHIPS and Science Act was the result of bipartisan negotiations and hundreds of discussions with industry experts, advocates, and officials—catalyzed by the need to stay technologically ahead of China. AI is generating similar momentum.
Schumer also sought to foster an innovation hub in his home state of New York through CHIPS, and he’s touting the same for AI. Schumer’s right-hand Republican on CHIPS was Young, whom he tapped for the AI group.
Close observers of the CHIPS negotiations lauded the uncommon pairing of Schumer and Young, saying a shared interest in preserving US competitiveness bolstered their dynamic.
“It’s sort of a logical extension,” said Margaret McCarthy, vice president of government affairs at the Information Technology Industry Council. Young played a “crucial role during the 11th hour” to beef up the CHIPS bill by securing investments in science and tech, she said.
Though parallels exist, addressing AI is more complicated than semiconductors, observers say. Lawmakers had been fine-tuning semiconductor legislation for years, but this explosion of AI “hit most of us like a big bang, right?” Young said. “And we knew that we were going to be looked to to respond in some way.”
Young said he was eager to jump into the AI conversation given the “connective tissue” he had with Schumer and tech leaders through their work on CHIPS. Young, who cosponsored an early piece of AI legislation in 2017, learned about its impact on warfare by meeting with the Marine Corps’ commandant around that same time.
Fast forward to today and “one of the scariest things I’ve learned is how this technology could be used to develop a wickedly lethal biological weapon—a superbug—conceivably,” Young, a Marine veteran, said. But that “is the very reason why we’re aiming to put a smart regulatory structure around it,” he added.
National Security Standpoint
Rounds likewise heard about AI’s advancements in defense for years. As the top Republican of the Senate Armed Services Committee’s cybersecurity panel since its creation in 2017 and a member of the AI caucus, Rounds heeded Schumer’s call to join the effort when AI catapulted into the national spotlight this spring.
“I said yes, because I think we’re going to accomplish a whole lot more,” the South Dakota Republican said.
Rounds is a natural fit for the bipartisan AI group, former Senate staffers said, not only because of his background in the subject, but because he brings a pragmatic and dedicated attitude to policymaking.
“He’s got a calming presence, and he’s genuinely a nice guy,” said Daniel Lerner, who previously served on the Armed Services Committee for almost a decade and worked closely with Rounds on the cyber panel.
“He’s really focused on understanding the issues and ultimately doing what’s right from a national security standpoint, as well as for the country,” Lerner said. Rounds would make himself available for hours to dig deep into problems, he added.
Rounds emphasized that Congress must also invest in AI to spur breakthroughs across sectors including healthcare, transportation, and energy. He witnessed AI’s potential firsthand when his wife, Jean Rounds, was undergoing AI-informed cancer treatment before she died in 2021.
“She had some treatments that literally were very, very successful and extended her life by a period of about five months,” Rounds said. “That told me that there was real hope for the use of some really exotic tools that two years earlier did not exist.”
Seizing This Moment
Rounds now co-chairs the Senate AI caucus with Heinrich, replacing Portman after he retired at the end of the last Congress.
Heinrich and Portman founded the caucus in 2019. Since then, the duo ballooned to 14 senators and the caucus promoted 15 bipartisan bills that became law, directed at improving the federal government’s deployment of AI, expanding research and development, and examining deepfake national security threats.
“Senator Heinrich deserves credit on the Democratic side for also being out there, you know, ahead of the game,” Portman said.
AI Threats Confront Eager Congress Grappling With Learning Curve
Heinrich, an engineer and outdoorsman hailing from a state that’s home to major national labs, has long been a staunch advocate for science and tech. His technical training and dogged curiosity, characterized as hallmarks of his legislative style, prove useful for AI policymaking, former Senate staffers said.
Chip Somodevilla/Getty Images
Sens. Martin Heinrich (D-N.M.) and Chuck Schumer (D-N.Y.) at the US Capitol in 2017.
“He’s not the strident, political language kind of guy,” Joe Britton, who previously served as Heinrich’s chief of staff, said. “He’s more about: well, what’s the real issue here and how can we solve it?”
Portman shared the sentiment. The Ohio Republican recalls Heinrich chatting him up on the Senate floor during votes about issues like artificial intelligence, “which was such a nice contrast to the partisanship that was going on all around us.”
Heinrich said he’s hoping to carry on that bipartisan approach as AI discussions continue on Capitol Hill. Heinrich, Young, and Rounds, with Sen. Cory Booker (D-N.J.), before the August recess introduced a bill (S. 2714, H.R.5077) that would create a shared AI infrastructure for researchers to have access to computational tools and educational resources.
AI Research Infrastructure Gains Bipartisan Support in Congress
“The four of us, and a lot of other people, are working well in this space right now,” the New Mexico Democrat said. “We should take this moment in time as an opportunity to figure out what we have a bipartisan consensus for and put that together in a package.”
To contact the reporter on this story: Oma Seddiq at oseddiq@bloombergindustry.com
To contact the editors responsible for this story: Michaela Ross at mross@bgov.com; Robin Meszoly at rmeszoly@bgov.com
