WASHINGTON (NEXSTAR) — Lawmakers are slowly ramping up efforts to regulate artificial intelligence.

Rep. Seth Moulton (D-Mass.) says as a matter of national security, the number one priority must be setting up guardrails within the military here and abroad.

“Once it’s out of the box, it’s hard to put it back in,” he said.

Moulton says he fears that the U.S. government isn’t prepared to protect the nations from dangers posed by AI.

“The risk is not from like-minded nations the risk is from adversaries like Russia like China will totally disregard the moral norms that we have for using weapons like this,” he said.

AI developers warn, left in the wrong hands, their tools pose an existential threat to humanity. Moulton says he told the Pentagon about these very risks three years ago.

“What we need is actually a Geneva Conventions type of an agreement that all nations, friend and foe alike, agree to,” he said.

On Thursday, the UK prime minister and President Biden signed an agreement pledging to step up.

“Our job as leaders is to ensure this technological revolution makes us more secure and not less,” UK Prime Minister Rishi Sunak said.

On Capitol Hill, Moulton says he’s pushing to add a provision to this year’s defense bill to establish AI guidelines within the U.S. military.

“We’ve got to get ahead of this and, candidly, the Pentagon is behind,” he said.

This week, Senate Majority Leader Chuck Schumer (D-N.Y.) announced three Senate briefings on AI over the summer. One focuses specifically on the military’s use of the technology.

Sen. Josh Hawley (R-Mo.) is introducing his own bill to reign in AI and says Congress must work quickly.

“I do hope the defense bill takes cognizant of what China’s trying to do militarily,” he said.