Senator eyes updating NDAA with AI use guidance

"For this next NDAA, I've been working with some of my colleagues already on this and how do we address this issue within Anthropic. Obviously, rules of engagement is something every military has, they change over time."

"I think it is a reasonable thing to expect from any contractor that there are things we should and shouldn't be doing within the U.S. military. We're not Russia, we're not China or North Korea. We have to have a certain level of standard. And I think those standards actually help us."

"Combat demands flexibility and speed, and said, 'We're going to have to make exceptions' to having a human in the loop for autonomous offensive systems. Contractors and agencies looking to use advanced sensitive technologies should be having discussions 'up front' on how both parties intend to use the technology."

Senator Mark Kelly and colleagues are examining how current legal frameworks govern advanced AI technologies in military operations. They are considering updates to the fiscal year 2026 National Defense Authorization Act to address AI usage following tensions between Anthropic and the Department of Defense. Kelly acknowledged that military combat requires flexibility and speed, necessitating exceptions to human-in-the-loop requirements for autonomous offensive systems. He emphasized that contractors and agencies using advanced sensitive technologies should discuss intended applications upfront. Kelly stressed that the U.S. military must maintain ethical standards distinguishing it from adversarial nations, arguing these standards strengthen military effectiveness and alliance relationships.

Read at Nextgov.com