In this YouTube commentary, TJ Walker discusses Paul Krugman's analysis of the Pentagon’s recent decision to label AI company Anthropic, creator of the Claude system, as a “supply chain risk” and to prohibit its use by military contractors. Krugman argues the designation deviates from the legal definition of supply‑chain risk—which normally concerns sabotage by foreign adversaries—and instead stems from Anthropic’s refusal to allow its technology to be used for fully autonomous weapons or mass domestic surveillance. Walker expands on Krugman's points, criticizing the Pentagon and specifically Pete Hegseth for politicizing procurement, suggesting the move is driven by ideological disagreement rather than security concerns. He likens the action to authoritarian tactics seen in Russia and Hungary, frames it as a DEI‑driven effort that punishes companies with differing political views, and warns that sidelining a superior technology like Claude could weaken national security. Walker also highlights the irony that critics of diversity policies claim merit should prevail, yet the Pentagon’s decision appears to prioritize political alignment over performance. Throughout, Walker recommends Krugman's Substack and Daily Misses for deeper analysis.
Chapters
Economist Paul Krugman says a new Pentagon decision about artificial intelligence raises serious questions about politics and national security. The Defense Department has labeled the AI company Anthropic, creator of the Claude system, a, quote, supply chain risk, and ordered military systems and contractors to stop using the technology. Krugman argues the designation appears unrelated to to the legal definition of supply chain risk, which normally refers to sabotage or subversion by foreign adversaries. Instead, he says the conflict stems from Anthropic's position that its AI should not be used for fully autonomous weapons or mass domestic surveillance. Krugman argues the move could weaken national security by blocking tools many defense experts considered highly effective, The irony, he says, is that critics of diversity policies often claim merit should come first, yet this decision may sideline a widely used technology because of politics rather than performance. Now, this is an absolutely fantastic analysis from Paul Krugman, which is one reason why I pay for his Substack. I would highly recommend Paul Krugman if you can only get one source of information beyond what you get here at Filtered, I would recommend that you read his Daily Misses because he always goes deeper. I'll admit, when I first saw this story, I thought, okay, the Pentagon has a point. They have the right to tell a military contractor that if you supply us with services, software, whatever, we get to use it any way we want, and therefore we're not gonna use Anthropic. I figure, okay, that's their right, but that's only part of the story. What Krugman points out is if they had stopped there, they'd be okay. They'd be on defensible ground. But by officially labeling Anthropic a supply chain risk, they're telling the whole world and the whole tech world and the Amazons of the world, you can't even do business with this company because we essentially think they're going to sabotage us. Well, there's been no credible evidence whatsoever that anyone at Anthropic was trying to sabotage the US military. The real reason is Pete Hegseth and his henchmen at the Defense Department don't like the political opinions of people who work at Anthropic. They don't like the fact that Anthropic says, we don't want you to use our technology to do mass surveillance of Americans, or to just create drones and have it go off and fight wars without human intervention because who knows, they could kill everyone. That's their right to have that opinion. It's Pete Hegseth's right to have a different opinion. But we're really in new territory when the government says, because you have an opinion different from ours, not only are you not allowed to work for us, but we're essentially going to try to put you out of business by saying no one who does work for the Defense Department can have any dealings with you. This is really getting very close to what happens in Putin's Russia. This is what happens in Orbán's Hungary, where you take some position that angers the dear leader and you've essentially destroyed your business. My friends, that's not free market capitalism. That is woke DEI run amok. It's just that in this case, the woke policies mean that unless you are a MAGA lover and support everything MAGA does, We're going to punish you. You're not allowed to have different opinions. In this case, what it means is we're going to have affirmative action for technology companies that are less effective than Claude because we like what they say. We like their political values more. This is supposedly everything that Hegseth and Trump and company are opposed. But here they are implementing their own essentially DEI program. It's a diversity program for technology companies that aren't as good as Claude. And by all accounts, Anthropic simply is better at what it does, which is why the Defense Department is using it. There's nothing preventing the Defense Department of saying, okay, we're going to use you for all of these services because you're the best at that. And for other services where we need other things, we'll use other tech companies. There's nothing wrong with that. But to try to put this company out of business because they are not right-wing woke is an absolute travesty. And it shows that when people talk about the problems with DEI in our culture, there is a real problem, but the problem is coming from within by the people who complain about it the most.
