Lawmakers are pushing to impose limits on the Pentagon’s use of artificial intelligence as the Defense Department races to integrate the technology into everything from intelligence analysis to battlefield operations.
An emerging version of the annual defense policy bill may include AI guardrails, including a ban on the Defense Department firing autonomous systems to kill without human authorization, to spy on Americans, and to launch nuclear weapons.
Democratic Sen. Elissa Slotkin told NOTUS on Tuesday that she expectations legislation she introduced in March to be incorporated into the base text of the 2027 National Defense Authorization Act. The Senate Armed Services Committee is set to mark up the bill next week.
“Base text means there’s [bipartisan] agreement,” the Michigan senator said. “It wouldn’t have made it in there if it were a controversial thing, so I took that as a real sign that people know we need some left and right limits — reasonable and not overly dictatorial limits.”
Trending
Sen. Kirsten Gillibrand introduced legislation along similar lines Tuesday that would bar the Defense Department from using AI to launch nuclear weapons, conduct domestic surveillance on Americans, or deploy most fully autonomous weapons.
“Right now, the Pentagon is moving toward deploying incredibly powerful AI technology without commonsense guardrails in place, which could have catastrophic consequences that make all of us less safe,” Gillibrand said in a statement about her bill, the Secure and Accountable Military AI Act. “We must act now — not to stifle technological progress, but to establish clear rules of the road that keep humans in charge and keep AI’s use in warfare smart and safe.”
President Donald Trump has pushed rapid adoption of AI across the federal government while backing away from some proposed safety measures. Trump on Tuesday signed an executive order that directs various federal entities to shore up cybersecurity for critical infrastructure. It asks AI companies to voluntarily submit their new models for a 30-day government review before releasing them to the public.
Defense Secretary Pete Hegseth, meanwhile, argues that national security is at stake in a push for the military to quickly adopt artificial intelligence and integrate it on classified networks. That effort spurred a fight with Anthropic, one of the developers of that technology, after it took a moral stand in favor of ethical limits.
“We will judge AI models on this standard alone; factually accurate, mission relevant, without ideological constraints that limit lawful military applications,” Hegseth said at a SpaceX event in January. “Department of War AI will not be woke. It will work for us. We’re building war-ready weapons and systems, not chatbots for an Ivy League faculty lounge.”
But there’s plenty of AI skepticism in and out of the military.
Pope Leo XIV recently called for governments to robustly regulate AI and declared that it was “not permissible” to entrust it with lethal decisions.
Vice President JD Vance, along similar lines, used remarks at the Air Force Academy last week to urge servicemembers not to concede their decision-making to AI.
“As AI transforms the battlefield in some ways positively, in some ways not, I ask that you be jealous and selfish about your role as the decision maker in warfare,” Vance said in a commencement address.
Some military leaders are wary of how artificial intelligence will be employed. During a recent conference, the chief of U.S. Special Operations Command, Adm. Frank Bradley, said recently that U.S. troops “have to be very careful” about using AI.
While AI could identify potential targets, he said, “We, as humans, have to have the confidence that … it’s going to deliver violence only where we intend it to be delivered.”
Gillibrand’s bill would put new limits on how the Pentagon can use artificial intelligence for “high-consequence” applications — nuclear, lethal targeting, domestic surveillance, cyber — and require senior Defense Department leaders to sign off first and Congress to be notified.
It would also establish department policy that AI supports but does not substitute for human judgment in decisions involving force, detention, domestic surveillance, or other high-consequence AI applications. The bill generally bars fully autonomous weapons but allows semiautonomous systems for missile defense and to track, identify, and prioritize targets.
Author
