Rep. Josh Gottheimer (D-NJ) said Wednesday that artificial intelligence models should undergo mandatory government review before release, rejecting both calls to halt AI development and arguments that companies should be left to police themselves.
Speaking on CNN News Central, Gottheimer took aim at Sen. Bernie Sanders (I-VT), accusing him of wanting to make AI disappear entirely. “Some people have suggested, like Bernie Sanders and others, that you just kind of make it go away, throw it in a box, and chuck it into the Hudson River and hope that you never have to see it again. That’s obviously not practical,” Gottheimer said.
Instead, Gottheimer argued for a system in which major AI developers would be required to submit their models to the federal government for a national security review before deployment.
“We need very strong mandatory guardrails in place for a review of these models, right? Before they go out in the wild, the major labs, the ones on the most cutting edge, have to come, I believe, and go to the government and get a national security review to protect our critical infrastructure, to protect Americans, to make sure that our cybersecurity is safe.”
He said that oversight cannot be left to the companies building the technology. “I believe that can’t be something that’s done on their own. They can’t self-police. The government should have mandatory requirements for that,” Gottheimer said.
Gottheimer also rejected the opposite position, that government should stay entirely out of AI regulation. “The other side is saying, hey, the government shouldn’t be involved in this, let us handle this ourselves. That’s also ridiculous, right?” he said.
Gottheimer framed his stance as the product of work on an AI commission, saying the goal has been finding “a smart way to do this, with strong guardrails that protect people and protect our security” while still allowing the U.S. to “compete and win” in AI development.
