00:00So I just want you to help me understand this.
00:02In February, on the eve of Operation Epic Fury,
00:05you publicly designated Anthropic as a supply chain risk.
00:10However, this week it's been reported that the White House is now helping agencies
00:14get around this decision to access Anthropics technology.
00:17So the administration cannot credibly make both claims simultaneously.
00:24So before I ask you about the inconsistency,
00:26I just want you to reconfirm what it is you plan to use this technology for.
00:32It's been publicly reported that the decision to label Anthropic a national security risk
00:36was influenced by your personal and very public contract dispute with Anthropic
00:41when the company said that its technology could not be used for fully autonomous weapons
00:46targeting or mass surveillance of Americans.
00:49So following up on your response to Senator Rounds earlier,
00:53can you confirm whether or not there will always be a human in the loop
00:58when AI is used for lethal targeting decisions?
01:03Well, first of all, on Anthropic, they would not agree to our terms of service.
01:07That would be like Boeing giving us airplanes and telling us who we can shoot at.
01:11This is not just about Anthropic, though.
01:13I just want to be clear.
01:15And also, Anthropic is run by an ideological lunatic who shouldn't have a...
01:19But that's not my question.
01:20My question is AI writ large.
01:21Decision-making over what we do.
01:23Writ large.
01:23We follow the law, Senator.
01:25Should AI...
01:25Will you confirm?
01:26You said this to Senator Rounds earlier, so I'm just asking for you to...
01:29We follow the law, but we don't have to sign a different terms of agreement with Anthropic.
01:33That's not...
01:33This is not the question.
01:35This is not about Anthropic.
01:36This is just an example.
01:37I want you to confirm that whether or not there will always be a human in the loop
01:43when AI is used in a kill chain for lethal targeting decisions,
01:47will there always be a human in a loop, or will AI make the decision?
01:51You said this to...
01:53We follow the law and humans make decisions.
01:58So you will confirm what you said to Senator Rounds, that a human will always be in the loop
02:03when AI is used.
02:05Which is why...
02:06I've just...
02:06That's part of our terms of service anyway.
02:10You know...
02:11It's how we operate.
02:12All I want to say is this.
02:14There is a DOD directive, 3000.09, which mandates that autonomous and semi-autonomous weapon systems
02:25be designed to allow commanders and operators to exercise appropriate levels of human judgment
02:30over use of force.
02:32That is in the DOD directive.
02:34That's why we follow the law, Senator.
02:36So the answer is yes, Mr. Secretary.
02:39The answer is we follow the law.
02:40Absolutely.
02:41I think this is more important than following the law.
02:44I think that people want to know that AI isn't going to make lethal decisions, and it
02:49is critically important.
02:49AI is not making lethal decisions.
02:52That's what we want to hear.
02:53We're going to follow on that one.
02:57Subscribe to One India and never miss an update.
03:02Download the One India app now.
Comments