Senate Majority Leader Charles E. Schumer said the first gathering of the world’s top artificial intelligence experts with senators produced a consensus that Congress needs to act to regulate the development of AI technologies, but there was no agreement on the correct steps. 

“First, I asked everyone in the room, ‘Is government needed to play a role in regulating AI?’ and every single person raised their hands even though they had diverse views,” Schumer, D-N.Y., told reporters Wednesday after meeting with more than 20 tech CEOs and civil rights advocates in the first of several planned AI insight forums.

But the majority leader said opinions diverged on ideas like a proposal from Sens. Richard Blumenthal, D-Conn., and Josh Hawley, R-Mo., calling for a new federal agency that would issue licenses for certain kinds of AI technologies.

“Some people mentioned licensing and testing and other ways of regulation … there were various suggestions as to how to do it, but no consensus emerged yet,” Schumer said. 

Schumer was accompanied by Sens. Martin Heinrich, D-N.M.; Todd Young, R-Ind.; and Mike Rounds, R-S.D., who are working with the majority leader to draw up proposals. 

Elon Musk, the CEO of Tesla and owner of X, formerly known as Twitter, who was one of the participants, told reporters after the closed-door morning discussion that regulations for AI were similar to the debate in the 1960s over seatbelts in cars. 

“Seatbelts were opposed by the auto industry for a very long time, even though data was very clear that they were saving lives,” Musk said. “We wouldn’t want to be in that situation where we are fighting regulations even though it’s a safety thing.” 

The intent of the discussions was not “jumping into the deep end and making rules,” Musk said. “You start with a group formed to create an insight, to understand the situation, then you have the proposed rule-making, then you get some objections from industry, and then ultimately you get a sort of consensus.” 

He was joined by other tech executives such as Meta CEO Mark Zuckerberg and Google CEO Sundar Pichai.

Young, who also spoke after the morning session, said that in the absence of any regulation “the incentives are aligned towards those who deploy models quickly to consumers,” referring to generative AI and large language models that could pose risks to individuals and countries at large.

But, Young said, some companies have refrained from releasing advanced AI models “until we can put together a regulatory atmosphere.” 

The participants agreed that U.S. government needs to “help to deal with what we call transformational innovation” using AI technologies, with one proposing a $32 billion fund for “the kind of stuff that maximizes the benefits of AI,” Schumer said. Microsoft co-founder Bill Gates, for example, suggested AI technologies could help feed hungry people around the world, while another participant said they can help with “curing cancer,” according to Schumer.

Rumman Chowdhury, CEO of Humane Intelligence, who participated in the discussion, said the $32 billion suggestion came from Eric Schmidt, the former CEO of Google who previously chaired the National Security Commission on Artificial Intelligence. She was one of several advocates looking to ensure tech advances don’t infringe on civil rights. Others in the room included representatives from industries impacted by AI, such as Charles Rivkin, chairman and CEO of the Motion Picture Association.

Schumer offered another briefing following the afternoon session, which was also behind closed doors, saying the discussions turned to AI applications relating to health care and potential displacement of workers.

The technology’s role in the 2024 election is likely to be one of the most critical areas in which Congress may propose guardrails, Schumer said. Senators discussed ways to promote so-called watermarking technology that can identify material generated by AI as one of the solutions to deal with dangers of fake ads or political messages, he said.