Loading summary
Julia Curley
Want to sneak peek every outcome digital
Alan Rosenstein
twinit and take the lead in a
Scott R. Anderson
world of what ifs to navigate the
Alan Rosenstein
complexities of change with ease.
Scott R. Anderson
Transform the everyday with Siemens
Michael's Party Shop Announcer
Is your kid's birthday coming up? Don't stress the prep. The party Shop at Michael's is your one stop shop for everything from Bluey to Rodeo. Transform your space into a birthday wonderland with dream worthy tablescapes and decor starting at $0.99 plus get free helium inflation on select balloon styles. Whether you're crafting a one of a kind bash or grabbing the essentials, Michaels has everything to make their big day legendary. Shop now and order ahead for delivery or in store pickup. Michaels everything to celebrate anything
Bill.com Advertiser
this episode is brought to you by Bill, the intelligence finance platform that helps businesses and accounting firms scale with proven results when you're growing a business, the stakes get higher. You can't afford infrastructure that breaks under pressure. If you care about security, reliability and scale, I want to let you in on a secret. Bill is the foundational software that nearly half a million businesses and 98 of the top 100 U.S. accounting firms use to automate back office workflows, add secure controls to payment processes and scale without increased overhead. With AI powered Accounts Payable automation, Bill erases the busy work from capturing invoices, routing approvals and processing payments, syncing seamlessly with the top accounting software platforms so your books are always accurate. But Bill isn't just accounts payable, it supports the full payments workflow. Bill has processed over $1 trillion in transact, leveraging that expertise to help you manage, move and maximize your finances. Stop the guesswork and start scaling with the proven Choice. Go to Bill.comProven to talk with a payments expert and get a $150 gift card as a thank you. That's Bill.comProven terms and conditions apply. See Offer page for details.
Scott R. Anderson
Alan said, you're back on I feel obligated to ask you how the weather is in Minnesota.
Alan Rosenstein
It was so nice for a solid week, but today it's rainy and kind of gloomy. But you know it's good for the plants. Which I think is the thing that as a Midwesterner you're like obligated to say every time it rains you have to say you have to intone, but the plants are happy so it's fine.
Scott R. Anderson
You just have to put a shiny gloss on whatever might be happening out there.
Tyler McBrien
Or is that from the farmer origins?
Alan Rosenstein
I think it's from the fact that Minnesotans are such extreme gardeners because it's cold and we only have a limited gardening season that you then have to just get very excited about anything that will make your plants happy. And I'm not the gardener in the family, but I am the manual labor for the gardener in the family, so that's my contribution to the cause.
Scott R. Anderson
I feel like Garrison Keillor would blame Calvinism, right? Wasn't that the usual stalking horse for most Midwestern norms?
Tyler McBrien
He's always blaming Calvinism.
Alan Rosenstein
Blaming Calvinism is always a. A fairly safe bet in Minnesota.
Scott R. Anderson
Hello, everyone, and welcome back to Rational Security, the show where we invite you to join members of the Lawfare team as we try to make sense of the week's big national security news stories. We have a full house this week because we have a lot of news percolating. We have a few people joining us for just one segment because it was a topic that we're hoping to get to discuss a few weeks ago, and now we actually get to discuss it at least a little bit, preliminarily. Joining me this week and thrilled to have everyone here, we are joined first by Ari Tapatabai, a former public service fellow, now contributing editor here at lafair. Ari, thank you for coming back for just the first segment. We're happy to have you for that far at least.
Ari Tapatabai
Always fun. Thanks for having me.
Scott R. Anderson
Wonderful. Also joining us is Tyler McBrien, who I believe I previously described as the third mast of rational Security because of his frequency here. Tyler, thank you for helping keep the ship afloat.
Tyler McBrien
Always a pleasure.
Scott R. Anderson
Also joining us is, I think, can only be conceived of as an old, broken, crappy mast that we broke off and discarded and now use mostly for fire firewood. Co host emeritus Lawfare research director, Alan Rosenstein. Alan, thank you for coming back on the podcast.
Alan Rosenstein
I like to think of myself as flotsam and jetsam that's been thrown over and is just kind of like barely clinging on to the rat sex ship as it steams through the water.
Scott R. Anderson
Honestly, you can tell a movie I've watched recently that basically I basically just watched Master and Commander again for the first time a couple weeks ago. And now I have all these nautical metaphors kind of in my brain for the first time.
Alan Rosenstein
My God, like, talk about it.
Scott R. Anderson
No, no, no. First time in a while I watched it that was like my dad, of course, showed me that movie, made me watch it 100 times when I was a teenager. Like, everybody so good. But I do, now that I'm a dad, I tune back in. I gotta say, pretty top, top notch. But joining us for non nautical metaphors, as far as I'm aware, although feel free if you're up for it, is none other than our latest public Service Fellow edition here at Lawfare, Julia Curley. Julia, thank you for joining us.
Julia Curley
Thank you, Scott. So I'm like a 10 year listener, first time caller, I guess to Rational Security. This is like I have briefed presidents in my life, vice presidents met foreign leaders, but this is the first time my mom gets to watch me work. So this is pretty terrifying.
Scott R. Anderson
There you go. We don't know how the hell you got our phone number, but we'll take it. We're glad. We're glad you called in.
Alan Rosenstein
And just to be, just to be clear, Julia, this is much, much, much higher stakes than all that.
Julia Curley
Absolutely. 100%.
Scott R. Anderson
Yeah, exactly. Exactly. No margin for error here. National security, actually, that's the whole point of the show. But that's okay. We are have a lot to get through this week and I want to get right into it. Topic one for this week, fission accomplished. After nearly four months of war, the United States and Iran have reached a deal to end the conflict, perhaps with Trump declaring it complete and authorizing the opening of the Strait of Hormuz ahead of a formal signing ceremony set for June 19th in Switzerland. We think maybe. But the agreement leaves enormous questions unresolved, from the fate of Iran's enriched uranium to sanctions relief to whether the ceasefire extends to Israel's campaign in Lebanon. Is this the durable peace that Trump and others claim or a fragile pause if that, papering over the hardest issues in the conflict? Topic 2 Model misbehavior days after Anthropic publicly released its powerful new Claude Fable 5 model, the Commerce Department imposed export controls barring any foreign national inside or outside the United States from accessing it, forcing the company to disable the model worldwide. The administration says Anthropic recklessly refused to fix a dangerous jailbreak. Anthropic says it was a narrow, non serious vulnerability and that the order reflects a misunderstanding. What does this episode tell us about the government's expanding use of export controls on AI and its increasingly adversarial relationship with one of the country's leading labs? Topic 3 Bad Vibrations in One of her final acts as director of National Intelligence, Tulsi Gabbard rescinded two Biden era intelligence assessments that had cast doubt on whether a foreign adversary was behind Havana Syndrome. The mysterious ailments afflicting U.S. intelligence personnel and diplomats around the world. Gabbard's office says the prior assessments cherry picked intelligence to support a predetermined conclusion, but critics worry about a politically motivated rewrite of analytical findings on the way out the door. What should we make of this last minute reversal and what does it mean for the future of the Vanna syndrome debate as well as Gabbard's legacy as dni? So our first topic is a topic that I originally had you on, Ari, a few weeks ago to talk about the first time when Trump told us and promised us that in fact a deal was about to be done. Then none came about and we had you on the show. We kind of talked about how there wasn't a deal. Now we're kind of the same position, I guess. We think the deal is about to be signed Friday. Both sides seem to agree something's going to get signed in Switzerland on Friday. Nobody's released it publicly yet. We've gotten a bunch of media reports about what it contains, although those have been almost uniformly contested by people in the administration, so we don't know how reliable they are. The one thing everybody seems to agree that it agrees on is that we'll have a 60 day pause in any sort of hostilities, extending the ceasefire, deepening the ceasefire, I guess presumably because it seems to suggest there'll be more than just the ceasefire in terms of a cessation of hostilities and that there will be some opening of the Strait of Hormuz, although it also seems to be suggested and most people seem to be conceding with some strong degree of Iranian and Omani, the two kind of countries bordering the Strait, control all other issues. The nuclear file, other files are getting kicked to that 60 day mark where they're going to have 660 window to try and reach negotiation on those. There may be other elements of the deal. We've seen media reports saying that sanctions relief is included, potentially substantial sanctions relief in even this first stage. We don't know yet. That hasn't been confirmed. Vice President J.D. vance has said no U.S. dollars are going to the Iranians. That doesn't mean that you're not going to see sanctions really allow Iranian dollars frozen to go back to Iran or portly Iranian dollars or other foreign money. We don't know exactly what these other yet. Presumably we'll find out at some point, but we're not 100% sure. It's been a few days now. We don't have any real details. So let me turn it over to you. We talked about the potential for an agreement, what a negotiated end of this conflict might look like on this podcast a dozen times at this point how does this fit in with your expectations, your concerns, your hopes? Where is this on the spectrum of pessimism to optimism? And what are you looking at? What are the big parts that stand out to you that are remarkable and worth noting on?
Ari Tapatabai
Yeah, I mean, I think it is essentially aligned with what we expected and unfortunately, because. And we should start with the premise that you mentioned at the outset, which is that we actually don't have the official text yet. What we do have is a bunch of pieces of the text and then a full text that was essentially posted by Iranian state media. The President said that's not the real text. This morning, Bloomberg and CNN and others have been posting essentially that text now saying that They've confirmed with US officials and other officials from the G7 countries who've been briefed on the deal that this is actually the text and it highlights all of the things that you talked about as a 14 point text that is actually very thin in terms of the details. My first reaction looking through the 14 points was, did a lawyer approve of this? Because let me say that from my experience and Scott, you'll notice too, and others will as well in this conversation. You know, when we're negotiating a nuclear agreement a few years ago, I do not think that our lawyers would have left us, you know, leave the room alive if we had walked away with this kind of text that leaves so much unsaid, that leaves so much gray, and so much room for maneuver and interpretation down the line. So that's the number one thing. Would love to hear your perspective as a, as an actual lawyer on some of the language that is used here. The second thing is that as we had expected, this deal really does focus on the deepening, if you will, or formalizing of the ceasefire that has been in place. It does not really touch on the issue of the nuclear program, which is why we're here to begin with. It kind of kicks that can down the road to the 60 days from now when we have an actual agreement. That is deeply problematic because we still have a significant amount of highly enriched uranium within Iran. The International Atomic Energy Agency does not have the continuity of knowledge that it would need to be able to verify that Iran's nuclear program is indeed peaceful. And the Trump administration does not seem particularly bothered by any of these details. The third thing that I think is really noteworthy here is that the deal seems to promise certain things on behalf of Israel, including the fact that Israel is going to essentially hold back and Israel has made it clear that it is not party to this deal, which it is not, and that it is not going to abide by this deal. And so the risk for Israel's operations in Lebanon or elsewhere, kind of, you know, derailing this whole process, are real. So back to where we were a couple of weeks ago when we last had this conversation. Yeah, okay. We now have more of a framework for, if you will. I've been calling it an interim deal because it's not a real comprehensive deal. We may have this interim deal, although even this morning the President was saying something to the effect of, you know, the texts that are floating around, like those are not real texts. We haven't really fully agreed to these things. So it's all very murky as to what's real, what's not, et cetera. But nonetheless, this interim deal just seems like a stepping stone to the next thing. It seems to really kick down the more challenging, more substantive, more complex issues till later on. And there are a number of spoilers and number of factors that can still essentially derail this whole process either between now and the 60 days from now or even after that. And that includes also, in addition to Israel, opposition from the President's own base. We know that a number of Republicans, and we've talked about this before on Rational Security, a number of Republicans, both in Congress and within his own administration, have for a long time opposed any negotiations with Iran. And then to, to kind of, you know, upfront, essentially tell Iran you're going to get billions and billions of dollars. Remember the pallets of cash from the Obama administration that was, you know, so controversial like this is essentially giving Iran a lot for not a whole lot, in return for essentially a vague commitment that they won't pursue a nuclear weapon, which, by the way, as far as I've seen, the draft doesn't really mention the Nuclear Non Proliferation Treaty, but they're already supposed to not be pursuing a nuclear weapon under the npt. So it's not really clear what we get on the nuclear front. It doesn't really touch on the proxies and missile programs. And I'm not even going to start talking about the human rights issue, which is its own bucket. And it's obviously not addressed in any of this. So in essence, the deal barely touches on any of the issues of concern that any Republicans or anyone who's kind of watching this space might have with Iran, which tells me that this is not a particularly. Even if we end up with a signing ceremony on Friday in Switzerland and it sets the clock for the 60 days and we end up with A deal. I don't know how sustainable this whole process is going to be and whether it's going to lead to viable agreement on the other end of it.
Alan Rosenstein
That was really helpful, Ari. But can I just ask you a much more basic question from the perspective of not an IR expert, just a guy who reads the New York Times? Did we lose the war in Iran? Because it really feels like we just lost the war in Iran. And I mean, you know, people lose war sometimes. But I don't know, it's hard for me to read this and not view this as anything but. We return to the status quo, except we drop all the sanctions against Iran and Iran now gets to skim money off the Strait of Hormuz. What am I missing?
Ari Tapatabai
No, I mean, you know, people have been describing it as the status quo. I think it's actually worse than the status quo because the status quo would imply that we did not just drop billions and billions of dollars to degrade very cheap Iranian capabilities that they can rebuild pretty quickly, pretty easily on the cheap. And by the way, we're actually apparently paying them to do so. It would imply that you don't have this kind of rift between the United States and Israel that is now more and more out in the open. It would imply that we don't have a regime that is more radical than it was three, four months ago. And look, the administration's talking point seems to be this morning on various news shows and so on, that we're really looking at Iran changing its behavior fundamentally. This is a very different regime than the one we had six months ago, and it's just not true. I mean, I've been raging on this, on rational security for months now.
Alan Rosenstein
Aren't the crazies running Iran even crazier now than the crazies that used to run Iran before we killed all them?
Ari Tapatabai
That is a very good way of putting it. Yes. You have a much more radical, more inexperienced crew there. And by the way, if you've been paying attention, we'll go back to human rights issue. If you, if anybody's been paying attention, Amnesty International just a couple of weeks ago talked about how this regime that we are now negotiating with and giving billions to is using the war as a cover to further continue to repress people that, you know, it's been essentially conducting mass arbitrary arrests. It's been intensifying repression. The executions, according to another human rights report just this week, are happening at a pace that has been unmatched in recent decades. So we are not dealing with a different reformed, enlightened regime. We're dealing with the same regime, but worse. So this not just the status quo ante, we've draw billions of dollars, deplete our stocks of munitions and gone damage to our bases and lost service members and by the way killed a bunch of people throughout the Middle east to end up with a worse regime than we had a few months ago. And that to me is so yes. The short answer to your question, Alan, is I think that this is a pretty epic and it's in the name already of the operation, but it's a pretty epic defeat, I think, for the United States.
Scott R. Anderson
Julia, I want to pull you in on this as well because I know there's an issue that you've worked on. I'll do a quick nod to Ari's because she invoked lawyers, which is always a mistake. So I'll do it. As a former State Department lawyer, I'll just note that generally an MOU is a term that's generally used for a non binding political commitment, not something legally binding. Doesn't have to be. There are MOUs that are considered legally binding. That's generally the assumption. And so it'll really depend on the verbiage because this is a draft. I suspect this may not have gone like a final scrub if it goes through a final scrub for between people who are familiar with U.S. agreement practice, because there is some weird verbiage that you would normally see State Department lawyers try and scrub out of this if they want it to be a non binding. Importantly, notably, I think they really need to argue not only that's a non binding, but to be careful on the nuclear file because if they touch too much on the nuclear file, I think there's a really strong argument Congress has to review this agreement under Inara. That's the law that it passed in 2015 to review the JCPOA, which involves, among other things, opportunities for Congress to disrupt this. Maybe that's not something the Trump administration is worried about given Republican control of both chambers, given the fact that I think everybody wants the Strait of Hormuz reopened. But it does complicate things, certainly on the timing front that they've laid out that they don't seem to be taking that into account here. So my guess is they're going to argue this doesn't have to go through Inara because it doesn't do anything on the nuclear file yet. That's going to come after 60 days. But there is a little tension there between some of the rhetoric the administration has used to describe this and the other outcomes of this sort of conflict. But Julia, let me come to you more on the policy substance for folks who haven't are familiar with Julia yet and she's relatively new with us. She is a veteran of the intelligence community and the National Security Council and has worked on issues related to Iran, the Middle east off among a variety of other issues over her time in government. Talk to us a little about your assessment of this, where this came out and how it fits into the broader strategic consideration that's gone around Iran, an issue set that presidents of both parties have really wrestled with for 20 plus not more than 20 years, but at least the 20 plus years around dynamics really similar to this through the JCPOA and then for the 10 years since that's been nixed, more or less. So where does this fit in on the range of potential outcomes there and how do you assess?
Julia Curley
Thanks, Scott. So first of all, since you mentioned my background, I have to say up front my views on this podcast are based on the open, unclassified record and of course do not necessarily represent those of the United States government. That all being said, I don't think we should mince words. The United States won every battle of this conflict. The US Military accorded itself as the finest fighting force in the world, but the United States lost this war. We lost 14 servicemen, we depleted critical munitions, we alienated our allies. We all paid the price at the pump and we showed ourselves impotent to reopen the strait. If you don't think people in Russia and China have observed this and drawn the conclusion that President Trump will back down in the face of serious political costs, or that the US Lacks the endurance for a protracted conflict, this has been a strategic disaster regardless of how the deal is eventually signed and worked out in Geneva.
Tyler McBrien
One thing that I've just been frustrated by and wondering about throughout this entire saga is this question of legitimacy. So in terms of my experience is trying to follow the negotiations every time Trump posts on Truth Social or gives an offhand comment to a reporter that a deal will be signed within a day, a deal will be signed with this provision and then it becomes like a New York Times news alert and then it doesn't materialize or it's just wildly misrepresented. And I found myself sort of of relying more on comments by the Iranian foreign minister as reliable, more so than the US President. But I was also thinking about the other side and maybe Ari, you are best positioned to answer this or Julia, of the legitimacy of this forthcoming deal in Iran and whether that's even, I don't know, the correct frame. I did hear one interview with an Iranian, I believe, official that it was portraying it as some betrayal of the regime and just the fact of the negotiations. So I don't know, is there even a way we can get a sense of the popular support of a deal, this deal negotiations in general?
Ari Tapatabai
I'll start. I'm really curious what Julia thinks, too. So, look, let's start with the premise that it's just difficult to know how people feel about this conflict. Just because I've said this before, it's hard to characterize the views of 93 million people, very diverse country. And also, by the way, we don't have great polling and we have even this great polling now. But the truth is that you've had, you know, 2026 has been a hard year if you're Iranian, right. You started with those mass protests that led to the greatest kind of, you know, repression that the country has had in decades and decades. Thousands killed, many more arrested. And, you know, there was this brief, I guess, moment of hope that maybe that would lead somewhere. And then on the back end of this, you end up in this conflict that kills again, thousands throughout the region, probably a few thousand at least in Iran, school children bombed all of those things and then leaves the Iranian people worse off, essentially because it has entrenched this regime and in some ways, I hate to say it legitimized. Like we are currently by way of this MOU and future agreement and with this administration saying again and again this is a different regime, we can live with this regime. The president having really nice words to say about the new supreme leader who is ordering said repression and who is green lighting the attacks on, you know, our bases and so on and so forth. Like there is a degree of legitimization that is happening that is just, you know, is just unfathomable to me. And I think that we would be having a very conversation, different conversation around it if this were not a Republican president, if it weren't Donald Trump specifically. So all of this to say that I assume that a lot of people in Iran do want this war to end because it is just, it's not a way to live. But I don't know how people feel about what comes next. And I think there's a sense of despair that at least anecdotally, people are capturing and the reporting coming out of Iran that this is again, that they're now, after weeks and thousands of deaths, people thought that maybe a reform of some kind. Maybe a change in regime in some kind was possible, that none of that is going to happen and are just going to continue with the same kind of infrastructure and the same regime structure. But just to Alan's point, crazier people in place. So. So yeah, but then to also pick up on your last piece there, Tyler, about some people have been opposing the deal. You'll always have some of that in that system, right? People will always come out and say, well, any deal is kind of the mirror of this side in a way. There are people who fundamentally do not want to negotiate with the United States and you'll always have that crew. But ultimately it doesn't matter. Right? The system is going to push in the direction it wants to push in. And right now it views this deal as putting it in a better place than it was a few months ago. And so it is going to agree to it.
Scott R. Anderson
Julia, how about you?
Julia Curley
Yeah, it does seem like the IRGC and the Basij are fully in control and nothing has weakened their grip on the population. You know, if you look historically at other conflicts, an air only campaign, if you look at Serbia and elsewhere, often tends to reinforce the regime and delegitimize the opposition who doesn't feel like they can come out in the street while their own country is being bombed from above. And so it'd be quite natural that this would reinforce the regime. And that's, I think, what we're seeing. I would also say, just in terms of the negotiating posture, Iran has observed the United States shift goalposts in this conflict several times. At one point, the President was declaring the need for Iran's unconditional surrender. And now we are basically trying to get to a ceasefire deal that would restore the status quo anti and perhaps accept a nuclear deal that's worse than what was on the table before we entered into this conflict. So from that perspective, negotiating with the United States, I don't think the Iranians are going to be in a hurry to reach a deal when they know we will continue to negotiate with ourselves. And they'll likely end up with a better deal if they just wait a little longer.
Scott R. Anderson
I think that's really the key nub of understanding the dynamics here. We talked on this podcast in the past weeks about the weird pressure valve of Hormuz. Right. For the Iranians and for the global community. The Iranians, if they keep it complete shut off, eventually that is going to push the rest of the international community to say we have to do something here, even if it is a use of force, even if it is something we don't necessarily agree with, it's not actually in Iran's interest to completely shut off the Strait of Hormuz. They need to release some of that tension. But if they release it enough that you can get the United States and the international community kind of accepting a new status quo, relying on it, all of a sudden, that gives you a political tool that's really valuable. And when you're not accepting that political tool, you have a new source of revenue. Revenue because one of the concessions they've extracted from this appears to be some ability to impose a fee structure on transit. Now, that is something that would raise real international law questions and set a very problematic precedent if it was structured as a passage fee. There are circumstances where you see countries impose things like services fee for piloting and things like that in particularly dangerous straits. I think Turkey does this in the Bosphorus and stuff like this, where there are some sort of fee structures and service fees that can be allowed by certain countries around straits, around international bodies of water. Maybe you can structure. Presumably this is going to be structured that way. There's a great discussion with, I think, the editor in chief of Lloyd's List, which is kind of like the leading publication on international shipping on Iran. The latest the Telegraphs Iran podcast just in their last episode. I highly recommend it. It goes through all the weird contours of the industry around the space, but it'll be really problematic precedent. But even if you can square it with international law, and there may be a way you can do that here to some extent, at least kind of paper over it, it's still another revenue stream that Iran, that they did not have prior to this. Prior to this, ships could pass through these waters relatively freely. And now Iran, if it wants to, can shut that off, that flow relatively easily. It can do it in a quiet passive way by saying, oh no, we can't give you the license, you need to pass this anymore, maybe because of some pretextual reason and begin to put pressure that way softly. Or they can do it harder by shutting it down. Or they can maybe able to reinstate a more selective basis for transiting things. And the real damage of this conflict, I think, is that the United States and international community has shown that it's not able or willing to force the straits open like I think it was always assumed they could or would if Iran pushed too hard against them. If you really want to know the extent to which the United States has been put in a difficult position, listen to that podcast I just mentioned, he describes how the United States Navy has basically been leading shadow fleets of vessels along the Omani side of the strait to get some traffic reflowing for the last several weeks. This is the kind of quiet follow on to the Project Freedom initiative that was abandoned a few days after, in early May, after the 60 day deadline passed. That some effort at version of that has proceeded quietly. But it requires all the vessels to turn off all their tracking devices, turn off their lights, go silent. It's not something you can do at a high volume and it's a very high risk maneuver. Otherwise they wouldn't have to be taking all these measures. That doesn't mean the United States couldn't do more. The international community couldn't do more. They could invade southern Iran and occupy it. Right. But the cost of that are so dramatic that, that their willingness to do that just isn't there. And I think it's shown Iran that it actually has a lot more leverage over Hormuz than it assumed it had going into this conflict. And there probably is going to be a higher willingness to use it, I suspect.
Ari Tapatabai
So, you know, I think to me one of the biggest costs of this conflict is that we have allowed Iran to validate its doctrine that it had been developing for decades. As you've been talking through the various kind of scenarios that existed for the Strait and what it could look like going forward, I think Iran had obviously been planning for this for a long time, but this conflict really allowed for it to refine that whole scenario planning and think through the costs and benefits and think through what implementation would actually look like. So one of the biggest things for me is that in these types of scenarios, if you're going to go in, you've got to go all in and you're going to have to kind of take care of the problem. And if not, all you're doing is you're allowing the adversary to learn a bunch of lessons. And it's not just Iran, it's also Russia and China, right, who are going to be studying this conflict, I suspect, and who are going to learn a bunch of lessons from how we fight, where we fight, what we do and don't have the appetite to do. And it doesn't help that you have a president who sits there and talks very plainly and openly about well, we don't have the appetite to do X, Y and Z. And you know, that was to Julian point, right. Of it's been made abundantly clear in this conflict for with no tangible results what our weaknesses are. And that to me is deeply problematic.
Scott R. Anderson
So, Julia, I want to come to you another part of this. In conversations I've had with other veterans of the intelligence community, they underscore that this outcome to some extent was always a point of concern that people have been watching the conference. I think, I don't think this is a secret. I think people who studied Iran, who looked at Iran from an academic perspective, from other policy perspectives, had a similar concern is that that if you upset, however problematic the status quo ante was in Iran, if you seriously upset it, you don't 100% know what's going to come next. And you have lots of equities that can be toppled in different way, from a collapse of the regime that could lead to all sorts of regional problems, to the assertion of a new, robust, more hard right wing regime, which may be what we now have in Tehran, to things like the Strait of Hormuz disrupting global traffic, which is probably the most immediate consequence. And the Trump administration actually still seems sensitive to this last year, in the last summer, when it did its first round of strikes on Iran, extremely limited, aimed at the nuclear program, and then immediately led to a cessation, a pressure on Israel to cease its broader military campaign. The rhetoric around that, Marco Rubio and other people said, is that we are worried about regime stability. Not quite in so many words, but that's how I've read it. And I still think I've read it. I went back and reviewed it this week because I was like, was I making up that they actually had that view and it was there, it's in the talking points. So that view really shifted, at least in the president. I mean, that's not surprising. We all know Trump is mercurial. We all know he had a direct line to Netanyahu, who reportedly could persuade him to take this action. And he tends to follow the advice of the last person he talked to. So maybe he was the last person he talked to before he made a decision. But there was no pushback. We have this account from, I think it was from Maggie Haberman about the decision to go to the war where it relays it says no cabinet official actually backed Trump's decision to do this. They just weren't willing to say no.
Michael's Party Shop Announcer
No.
Scott R. Anderson
And then it kind of celebrates JD Vance as the person who got closest to saying no without actually saying it. What does that tell us about the state of decision making in the heart of this administration and what are the factors that contributed to that?
Julia Curley
I think it's fairly obvious that it's been uninformed. The Iranian ability to close the Strait of Hormuz, the regime's ability to survive the death of the Supreme Leader. Most of the problems that we've been talking about in this podcast were foreseeable and have been foreseen in innumerable war games and assessments that have been made public over the years. But we should think about what occurred in the run up to this conflict. The administration, in several public moves, clearly sent a message to the intelligence community that dissent would be punished and punished with the end of a career stripping of clearances and the public firing of senior officials, such as the director of the Defense Intelligence Agency, who was fired after his battle. Damage assessments that his agency produced publicly contradicted the president. Analysis at every phase of this was shut out, punished, or ignored. So this is the price of politicized intelligence, so that the US Is blundering into foreseeable crises that no one is willing to forecast.
Alan Rosenstein
So before we close on this, I do want to talk about Israel's position in all of this again, and I'd like someone to explain this to me because I am not the IR professional in the room. This seems very bad for them. Off the top of my head, it seems like Trump is going to sell them out, which again, seemed obvious in retrospect that he was going to sell them out the moment this got difficult. And also now Israel's standing in the international community, already somewhat tenuous, is that much worse because Israel contributed, and we can debate how much so. But clearly they were a motivating factor in all of this, in something that was actually quite disruptive to the global economy. And then in the United States, I mean, the amount of anti Semitism that this is going to provoke on the left, on the right, God knows where, and support the idea that, you know, American foreign policy is run by shadowy Jews. I mean, seems really profound because in this case there does seem to be at least a kernel of truth to the idea that America got into this war in part because Israel pushed it into it. Which of course doesn't justify anti Semitism. But again, it's not a great data point. So again, this seems bad for the Jews, which again, is not the only question anyone should care about. But to the extent you care about it, this seems like a disaster. Again, am I wrong or is this really, really bad?
Scott R. Anderson
I would say it's probably not great. Yeah. I did a long conversation with Joel Braunold about this last week for the Law Firm podcast. I strongly recommend folks check that out for Lawford Daily. And what we talked about this is that it's not just the decision to go to war with Iran. That was definitely Netanyahu was kind of the indispensable element of that in his conversation, at least if we take the Maggie Haberman account accurately of that kind of initial decision. We then also had the decision of Israel to open another front with Lebanon, where they did see some provocations. There were some rockets shot by Hezbollah, but they responded with a very aggressive military campaign that has them now occupying big swaths of Lebanon south of Latani. And you are in a situation where the domestic pressure on Netanyahu is we've got to find a way to resolve this conflict with Hezbollah because northern Israel is under a constant state of threat, as they kind of have been for a year, multiple years at this point, really kind of since shortly after October 7th led to the creation of a Lebanon front off and on. And they want a solution to this. And if you can't get a military solution, and it's not clear what a military solution looks like, if you're still occupying southern Lebanon, something that was is a big cultural point of controversy and touchstone for the Israelis from their last very harmful and poorly remembered experience occupying big swaths of Lebanon, you can't get the military solution there, then you need a diplomatic one. And here we're in a situation where Netanyahu has put him itself in a situation where Lebanon is being dictated by the terms of the Iran war by the United States. He pretty clearly now cannot drift too far from following the United States prescription in regard to Lebanon, as that is in turn being driven by Iran's demands in the Iran US Negotiations without really compromising the US Israeli relationship. When you have President Trump openly rebuking Bibi Netanyahu, saying, I think he's crazy and he's lost it on these particular points and he owes me for keeping him out of jail. That's not a great sign about the state of the relationship. And Bibi went all in on the relationship with Trump in a way Israel has over the last several years that has big ramifications. We already know the Democratic Party views of the US Relationship with Israel has been shifting in a way that's still strong in a lot of fronts, but also critical on others, and sees a real problem, particularly in the departure from the two state solution. And now we've got this wing of the Republican Party that's openly critical of the US Relationship with Israel, and that includes, frankly, people with a lot of connections with J.T. vance, a potential future leader of the party, and a lot of people around Trump who have a line and may be able to inform that views moving forward. And when you consider that, for instance, things like the United States, States and Israel are up to renegotiate their long standing security assistance agreement under this administration, given that the centrality of the United States to defending Israel from rockets of various stripes, from Iran, from Hezbollah, from others over the last several months and years, there's a lot of pressure points that can be brought to bear. All of a sudden, Bibi's gone from having Trump as his closest ally to somebody who could actually use that leverage against him. I think that has a lot of Israelis pretty upset, and that's why you've seen really strong reactions against this by the Israelis and attacks by that. People are going to be challenging Bibi in elections later this year, including mostly from the right. Honestly, a lot of the rhetoric has been really focused on saying Bibi's not being hard enough on this sort of issue. But they're capitalized on the fact that he hasn't left himself space to maneuver at this point because he's pushed everything so far and so aggressively that there's just nowhere else he can really go without really risking this US Relationship. Is that ungenerous? Julia, do you agree or disagree? I know somebody, I know who's followed this stuff.
Julia Curley
No, I think all of that is right. And I think it also points to the. The lack of durability of the ceasefire if Israel is a co belligerent in the conflict, but is not a party to the ceasefire. And Netanyahu is under continual pressure to strike targets of opportunity in Lebanon. So whatever influence the President has over our Israeli partners, it's not perfect. He will respond to his domestic political requirements and this will be a persistent problem as we try to keep the ceasefire alive.
Scott R. Anderson
Yeah, yeah, yeah. It's a fair point. I mean, that's the big spoiler point, I think, is Lebanon and how committed Iran and the United States are to their partners there, respectively. We'll have to wait and see.
Bill.com Advertiser
Have you ever thought I really should be doing something to protect myself from stalkers, scammers and hackers? But you're not sure what. But I already took care of myself in this regard. And I'm going to tell you how you can do it, too. Here's what you do. You go to www.joindeleteme.com and you enter the promo code lawfare20 and you will get 20% off Delete Me. Delete Me removes your personal information that's being sold online. In the age of AI, we're all vulnerable to scammers using our personal data. And it's just floating out there in the Internet and they're sucking it up and they're using it against you.
Ari Tapatabai
You.
Bill.com Advertiser
You don't believe me? Google yourself and you're going to find your home address, your phone number, the name of family members. It's really unsettling. But the good news is Delete Me can help. It was named the top pick for data removal services by wirecutters, and it has never been more affordable. Lawfare listeners can get 20% off at www.joindelete deleteme.com lawfare20. With the code lawfare20, you can get an individual plan for a little over $8 a month. I use Delete Me. I used Delete Me before they started advertising with Lawfare. I actually endorse this product wholly sincerely. I have an online presence, and there are people out there that. That don't like me. And this helps take control of your data and keep your private life private. By signing up for Delete Me now at a special discount for our listeners. Get 20% off your delete me plan. When you go to www.joindeleteme.com lawfare20 and use the promo code lawfare20 at checkout. The only way to get 20% off is to go to www.joindeLeteme.com lawfare20 and enter code lawfare20 at checkout. That's www.joindeleteme dot com lawfare20 code lawfare20. You want to know what part of running lawfare I really hate? I mean, there's the parts that I love. I love the editorial process. I love writing, but there's this thing that I don't love, and it's actually running a business everybody's trying to run. Leaner, tighter budgets, smaller teams, higher expectations. We all want to be super productive. And the last thing I have time to think about is, you know, doing a manual payroll or chasing down an HR form. And I'll tell you what I wish I had known earlier that there was this thing called Gusto. And it is how small business owners get back time when every hour counts and when they'd rather be, you know, running lawfare. Gusto is online payroll and benefit software built for small businesses and small nonprofit organizations. It's all in one remote, friendly, and incredibly easy to use, so you can pay, hire onboard, and support your team from anywhere. It's automatic payroll tax filing, simple direct deposit, health benefits, commuter benefits, workman's comp, 401k, you name it. Gusto makes it simple and it has options for nearly every budget. It's ranked number one on G2's highest satisfaction products list for 2026. It's trusted by more than 400,000 small businesses and small nonprofit organizations. So try gusto today@gusto.com Lawfare and get three months free when you run your first payroll. That's three months of free payroll@gusto.com LawFare One more time Gusto.com Lawfare
Scott R. Anderson
this podcast
Alan Rosenstein
is brought to you by Thomson Reuters.
Scott R. Anderson
The best don't just do their work. They change what's possible. Cases won, audits completed, jobs saved. Behind every one of those moments is a professional who needed to get it right and did. Thomson Reuters builds the technology that sharpens insights, speeds up decision making, and powers the outcomes that matter. So when professionals act, the impact is felt by everyone. Be a changemaker visit tr.com changemakers
Alan Rosenstein
I'm
Scott R. Anderson
Jake Stauch, co founder and CEO of Serval. We built Serval to automate the IT work that slows companies down. Onboarding password resets, access to applications. My laptop stopped working. While employees wait for help, their real work is put on hold. IT desperately wants to automate this work and that's why they need Serval. You just tell Serval what you want to automate in plain English and it's built. No drag and drop workflows, no expensive consultants. Employees get unblocked and IT teams go from drowning in tickets to building what actually matters. With Cervel, it becomes the AI engine powering the entire company. This is a new way to run it. We guarantee you'll automate 50% of all tickets and we'll prove it to you in a free four week pilot. Go to servil.com acasta that's S E-R-V-A L.com acast is your kid's birthday coming up?
Michael's Party Shop Announcer
Don't stress the prep. The party Shop at Michael's is your one stop shop for everything from bluey to rodeo. Transform your space into a birthday wonderland with dreamworthy tablescapes and decor starting at $0.99 plus get free helium inflation on select balloon styles. Whether you're crafting a one of a kind bath cash or grabbing the essentials, Michaels has everything to make their big day legendary. Shop now and order ahead for delivery or in store pickup. Michaels everything to celebrate anything.
Scott R. Anderson
So we are going to come back and talk about this deal, I have no doubt, once we see a final draft 60 days from now when we see actual acts on the nuclear file. But we have other big national security news stories we need to spend some time on this week. So let us shift our attention back to the interwebs, I guess, somewhere, wherever geographically, they may lie.
Alan Rosenstein
The machine God, Scott.
Scott R. Anderson
The machine God.
Alan Rosenstein
Her term is the machine God.
Scott R. Anderson
Where we have a really pretty dramatic development happen on Friday where we saw the Fable model, which is the kind of consumer facing, somewhat sanitized, guardrailed version of the borderline mythic, Mythos model that Anthropic has developed that supposedly brings huge, huge dividends in terms of particularly cybersecurity, in terms of how it's applied, both in terms of identifying vulnerabilities and potentially exploiting them if it were to get into the wrong hands. We understand that on Friday evening, the Commerce Department gave a letter to Anthropic indicating that it was subjecting Fable to an export license or export control that would require licenses for any transfer of that technology, not just to foreign countries, but also to foreign nationals, even in the United States. States. That in turn, according to Anthropic, or I think has been strongly suggested by people around Anthropic, at least made it impossible for them to actually work or offer Fable because so many of their employees are foreign nationals and there's no way to identify the nationality of users, at least not currently built in the system. So they ended up pulling Fable offline just a few days, I think, three or four days after they released it. And now it's become this new chapter in this tension between the Trump administration and Anthropic that we saw going all the way back to efforts to put supply chain threat risks and other sorts of labels on Anthropic that are currently being challenged in court. Alan, you've been following this closer than just about anyone. You've written about it for us, for Lawfare. Talk to us about this latest action, where it fits into this broader story and what we should be thinking about making sense of all of this.
Alan Rosenstein
Yeah, so this is a very, very big deal. I mean, this is the beginning of essentially a de facto licensing regime for frontier models where if you're anthropic or presumably OpenAI or Google or whatever, you need to go to the government and ask permission before you can release one of these powerful models, because if you don't, the government can slap export controls on them. It's notable that this comes. It feels like it's like weeks and weeks but this all happens every three days. There's a new thing. This happened literally a few days or a few weeks after the administration released its AI Cybersecurity Executive order, which was notable for being extremely light touch and hands off and merely allowing for a voluntary. It started off as a 90 day period, but that was too much. So then they cut it down to a 30 day period where the United States government would be able to get access to some of these models to analyze them for cybersecurity risk. That EO was explicit. It literally said this is not a licensing regime. And while it is true that it is not a licensing regime, then a week later the United States government acts in a way that creates a de facto licensing regiment. So all of this is to say that the administration's policy towards AI is, as with so many things, not super coherent and rapidly changing. So I wrote about this for lawfare on Monday and then after I wrote about it, the letter from Lutnick, the Commerce letter, leaked or was disclosed. Now what's interesting, having had a chance to read the letter, is that it's actually not clear to me what, what commerce has export controlled. So if you read the letter very carefully, right, and you actually trace the authorities back to export control law, what Commerce is doing at least is it's exporting, export controlling the model, right? That's what they say. We are export controlling the model. But what is the model? Well, ultimately the model is a, I don't know, multi gigabyte file of weights of parameters. Right. You can think of it as sort of vaguely comparable to a compiled binary for traditional software. So that would prevent Anthropic from taking the model and putting it on a server outside the United States. Okay. And it would also prevent Anthropic from allowing any of its foreign employees, so non US citizen and non lawful permanent resident, non green card holders from accessing the actual model weights itself and working on it. That would be a big deal because I assume that at least some of Anthropic's employees are here on, you know, just temporary employment visas. And if that were the case, that would have a profound effect on the ability of frontier AI companies to get the best talent. But the letter does not on its face apply to model access, which is to say to accessing the model through the Anthropic API or through the Anthropic web chat interface. Now that's important because it confirms something that I and other export control dorks who have been following this have thought, which is that it's not actually clear that the experts. Export control laws apply to accessing US Software when it's cloud hosted. Bis. The. Oh God, I'm so bad at these acronyms. The Bureau of Industry and Security, which is the part of.
Scott R. Anderson
Industrial Security.
Alan Rosenstein
Yeah, sorry. Bureau of Industrial Security, thank you. That administers the commerce export controls has actually long maintained, and there are public legal opinions from BIS to this, to this effect, that access to American software as a service is not export controlled. And the House representatives actually earlier this year passed a law that would explicitly expand the scope of export controls to apply to remote access to American services. Now I think it actually makes sense for export control to apply to this. Right. I mean, if you're going to have export controls on US Technology, I think. I think probably API access should apply to it. But it's notable that the House felt the need and the Senate is now currently considering the need to expand the law because it suggests that the current law does not apply to this. So it's actually quite possible that Anthropic is not legally obligated to do what it did, which is to say to legally obligate it to block API access to everyone. Now, what's interesting then is that that tells me that unlike with the supply chain designation, which Anthropic. Anthropic is fighting in court and is currently winning. Ish. In its various litigation, Anthropic decided to not fight this even though it could have. Which tells me, I mean, that's not a legal decision so much as that's a strategic business decision. And this happens all the time. Right. Companies routinely will not fight things that they could otherwise fight. But it's notable to me that the letter does not appear to do what the Commerce Department is saying it does. And I can't tell if that's a combination of actual legal limitations and the Commerce Department is bluffing or. To piggyback on Julia's point from the earlier segment about the hollowing out of the United States national security bureaucracy, whether the people doing this are just not that good at their jobs, which is always a plausible possibility with this administration.
Scott R. Anderson
Yeah, I think the legal challenge angle and practice here is interesting. I have a little bit of experience with this that I think is relevant, although I won't pretend to be a deep export controls expert, particularly in the AI space. So some of this may be fudged a little bit.
Bill.com Advertiser
But.
Scott R. Anderson
But I think you're right. My understanding is that there has been a policy determination from the Commerce Department, fairly long standing at this point, that Cloud Services aren't something that we view as subject to export controls. That's an agency position. It's like a policy guidance. It's something they can change. I don't think it's necessarily prohibitive. Although you raise questions about potentially APA challenges. I'll get to those. Why those aren't always available in a second. In other contexts, the export controls are applied to services that are provided. So if you are a Boeing or Lockheed and you're. I'm just saying those companies randomly actually know what they're subject to. But let's say they were like advising a foreign company, a foreign government or a foreign entity on designing defense tech of some sort or maintaining defense tech. Right. Export controls can restrict that sort of activity. So I don't know if it's necessary and I don't think that's actually specifically based on the like movement of physical people overseas. I think it also has generally been applied to things like remote service. Right. At this point that's probably common a lot of these cases. So I'm not sure that this would be entirely novel, at least outside of kind of the cloud service context. The real issue of this here is that there's a lot of discretion built around this and that the way Congress has designed the ecra, it really makes it hard to legally challenge this because the ECRA strips a bunch of jurisdiction and makes it not makes different licensing determinations not subject to APA challenges as readily. So you have to find either a constitutional hook or Altavira's challenge or other sort of like higher level challenges. You can't just make as easily a arbitrary and capricious argument. I think. Does that seem right to you, Alan,
Alan Rosenstein
or am I off on this? So I'm not sure. And I should say I'm also not. I was, you know, many, many years ago when I was a young and innocent DOJ attorney, I was briefly one. A part of one of these export control sub IPCs. And I remember it being the most complicated thing I have ever experienced in my entire life. But based on that I will use, I will claim stolen valor as an export control expert. The question of providing services. Right. Is. Is an interesting one. I think the way, the classic way that the government would try to export control this if it wanted to is it would export the actual control technology information that comes out of the model. Right. So there is an argument that the government can export and control technology. And technology in the statutes and regulations is defined as including information necessary to operate other things that are export controlled and things that are definitely export controlled, are, for example, cyber intrusion software. So if the government is concerned that, and this does seem to be the nature of the concern, that you can use Fable to get information that would allow you to do cyber intrusions. What they could have done was export control the output from Fable. That gets that, but they didn't do that. Like, what's weird about the letter, right, is that the letter only export controls the model. And I'm pretty sure that that means the model waits, and it cannot simply mean all access to the model. So I'm not saying the government can't export control these models in a way that would require Anthropic to shut them down for everyone, but I'm not sure they have. And again, this is just like a basic competence question. And as to your point, Scott, of the difficulty of challenging this, yes, I agree it's very difficult to challenge the determinations themselves, but this would, in fact, be an ultravirus case. Right. Like, what Anthropic would go into court and say is not, hey, we think the Commerce Department is acting unreasonably or arbitrary or capriciously in export controlling our model, but not export controlling OpenAI's model, which has similar capabilities. It's that they're literally trying to do something that the statute does not allow them to do. I don't see anything that would prevent Anthropic from making that argument. Now, again, whether they want to. Whether the C Suite has decided they want to do that is a different question. They clearly have decided that they don't, at least not while negotiations with DC are ongoing. But it's just, again, like, this is. This is always the problem, right, with Trump stuff like, you want to spend the time debating the merits, but then there's, like, basic housekeeping of like. But did you even cite the right thing that you end up getting sidetracked into?
Tyler McBrien
I realize that the crux of this has to do with this flaw in the model that supposedly allows it to be jailbroken. And there's conflicting reports of this, of course, as we've been talking about. But I wanted to back up and talk a bit about the capabilities here and the power of the model, which are, I think, very cleverly marketed with the names of Fable and Mythos as a starting point. Did anyone here get to use Fable in the fleeting moment that it was in the hands of the consumer, and if so, whether you would have done things differently had you known it would be so summarily ripped from your hands shortly thereafter?
Alan Rosenstein
Yeah, I used Fable for, like, the few days that I could. And it was incredible. I mean I found it super, super helpful. Now to be clear, I'm not a nation state hacker trying to penetrate secure systems. So I can't speak to whether Fable is a massive uplift. And I also didn't try, right? But for my stupid vibe coding projects and my, I occasionally talk, I like to talk to Claude about interesting legal issues. To me, Fable seemed a lot smarter, right? It did seem like a step change in capabilities. And I will say that the kind of AI experts that I read who are often very skeptical of puffery from the labs also seem to think that Fable was a substantial step change in capability. So yeah, I don't think Fable is just marketing nonsense. I really think this is the next frontier. And this is important because without getting too much into the weeds of this, what it seems that Fable is, is, is the next size of pre training run from Anthropic, right? So basically the way these models are developed is you do a giant pre training run which is basically you take the entire Internet and now it's the entire Internet plus lots and lots and lots and lots of data sets and you train a base model to do basically next token prediction. And then once you have that, and that takes months and months to run, that's a huge training run, then what you'll do is then you'll do sort of fine tunings of that model or sort of post training runs based on, on other stuff that you want to get. And so what Fable seems to be is the next step in like you add a zero to the size of the pre training run and it's notable that that did lead to a substantial uplift in capability. And the reason for that is that an additional pre training run or adding a zero to the pre training run does not require any clever architecture, right? It does not require cleverness, it requires just a lot of compute. So what that tells us is that the scaling laws of capability are still alive and well. And if Anthropic can do this with Fable, that means that OpenAI GPT6 or whatever they call the next thing is going to be this level of model. XAI is going to get there. Google is going to get there with Gemini 3.5 Pro or Gemini 4 Meta is going to get there. The open source models are going to get there in six to nine months, the Chinese open source models. So if in fact Fable is the sort of model that triggers emergency national security freakout, that's not an anthropic only problem. That means that basically every single model from here on to forever is going to be considered this dangerous. And again, maybe it should be. But that obviously raises kind of. I mean, that just underscores the need for us to do. If we're going to have a licensing system, it's got to be not this. Like, it's got to be better thought out than whatever this bullet is.
Tyler McBrien
Talk about clever marketing. I see how you slid in the name of your other podcast scaling laws in there. I saw what you did there.
Scott R. Anderson
Hey.
Julia Curley
Hey.
Alan Rosenstein
Available on a podcast channel near you.
Scott R. Anderson
I mean, I think that gets kind of at the dynamic here, Alan, because I'm torn on this issue generally, I will say, because I think it actually, from a policy perspective, it kind of makes sense that you would want the US Government to be able to say, hey, a given company, you've got a model that is doing new things. It's a little premature. We're worried about safety. We need to pull back on that. And people agree with me. Me, Dario Amodai. Right. He says as much in essay he released, I think, just in the last week or two, kind of suggesting he's been a big fan of export controls in the chip context and the semiconductor context for reining in giving democratic countries an edge over China and authoritarian countries in developing AI models, which he thinks is a big challenge in this technology field. Right. And this could fit into that. You could see a situation where like, yeah, okay, this is gonna be a problem we're all gonna have to deal with in nine months. But damn, those nine months make a big difference because nine more months that we can use Mythos to get our system secured. So when Deepseek or something else comes up, we're going to be in a much more stable position than we would be otherwise. So even the timing thing, I don't think necessarily wears against this. The problem with this is two things. One, the competence problem. Because if your concern really is that the anthropic hasn't put the guardrails in place, disabling their ability to even use their internal employees to address that quickly seems like a problem if that's what you're designing. The foreign national in the United States element of this particularly seems problematic. Maybe there was some additional licenses they included in this. Maybe their intention was, we're going to immediately issue a license for your employees that are essential to fix this problem, and that just hasn't been reported. It's possible, I think it would be reported if it were actually happening. That just seems like an odd move. And Then more fundamentally, there's just a trust problem. We're in this situation where our priors are, the Trump administration's got a beef out for Anthropic, because it clearly does, because we know it's pursuing all sorts of other very legally questionable actions against Synthropic in other contexts. That may not be what have led to this. I actually think there's real reason to think maybe this is a good faith effort by the Trump administration to take a policy choice, perhaps misguided, perhaps poorly executed, but still acting in good faith. But how do you weigh that? How do you know we're all dealing with these black box information? You need some sort of regular process, governmental regulation to weigh all this. That's got credibility and neutrality. And you're really hard to get that these days. And we have all priors about Anthropic too. Like, a lot of people just trust Anthropic. So their questions about safety, because they've been on the head, on the safety curve and the rhetoric for a long time, gives them, I think, a lot of credibility with certain audiences. But imagine this for flipped. Imagine this for Xai. And we were, Elon Musk, like, rushing this model of saying, we've done this. We've done all these internal testing. We've been doing it for 60 days. Even though this new discretionary process you set up by your EO is supposed to be discretionary. We've done more than that. We're confident with that.
Michael's Party Shop Announcer
That.
Scott R. Anderson
Would you trust Elon Musk the same way you trust Anthropic? I would. I would not. I would say, and I think I have good reasons for that, but that's my own bias. Like, that's my own lens. Right. I know there are other people who don't agree with me the same way how you impose these sorts of frameworks. Like you need to have some sort of institutional arrangement other than executive discretion if you get there. And I just don't see institutionally Congress taking this on to do this anytime soon. I don't know what else kind of arrives other than a very messy debate that plays out in politics and policy realms around these issues as they kind of arise.
Alan Rosenstein
So, look, I agree with you 100%. I don't know. I wouldn't actually count Congress out on this one. I think Congress is actually taking these issues around AI regulation quite seriously. I mean, I have not dug into this, but my understanding is that the current drafts of the ndaa, the National Defense Authorization act, are actually going to try to limit the supply Chain risk authority that the government is trying to use against anthropic. Again, I haven't read the details. I don't know if it's good, et cetera, et cetera, but I actually think that this is something that Congress does care about and should care about. And maybe if there's a silver lining here, it's that this might actually be the thing that kicks Congress into gear to, you know, put some processes around this. But, yes, I mean, there has to be. I think it's obvious we're moving to a licensing regime. We already have a de facto licensing regime. Let's just everyone commit to the bit and start figuring out, okay, what is going to be the nature of the licensing regime, how can we make it limited, how can we make it actually expertise based? How do we avoid the First Amendment issues that are kind of lurking around? But yeah, I think it's clear that we have now entered into the second. I think the supply chain designation was kind of always, to me, the sign that we entered into the second phase of AI policymaking in this country, which was, okay, we're going to nationalize this to some extent. And I think the export control fight we're having now is just clearly like, okay, we're really in this phase. That's what we're talking about for the next year, everyone.
Scott R. Anderson
Yeah. And I think part of that is because this strikes me as a much bigger step in that direction because these are actually much more credible than the supply chain designation, 100%. I think there's much more stronger legal ground for the administration doing, even if it's unwise, it's just harder to challenge. And everybody understands there's some need for some sort of tool here. We just don't have it. But yeah, hopefully you're right. Hopefully this is the thing that maybe gets Congress moving on this particular front. For now, we have to move on to a different topic because one more topic want to get to Today we saw a dramatic moves by outgoing Director of National Intelligence Tulsi Gabbard, who has been notable, we all know from years and years prior to ceding to that office for being somebody who buys into what many view as conspiracy theories, particularly regarding US Activities overseas, sometimes with close dovetails with rhetorical lines advanced by Russian propaganda machines in Russian intelligence and other avenues. And we're seeing that inclination perhaps come back to the fore. In the last few weeks into office, we have perhaps most notably the repeal of a report by the Biden administration, an intelligence assessment that essentially said we cannot reach the Conclusion at this point that Havana Syndrome, this phenomena that's been affecting U.S. diplomats and U.S. intelligence officers overseas, many of them very tragically having really lifelong medical consequences from it or from whatever the variety of things that might be contributing to it. Where many people have said there's reason to think this is the result of some sort of energy directed energy weapon being implemented by Russia or someone else, the Biden administration or under the Biden administration, I should say the intelligence community concluded there's not much we can say. We can't affirmatively conclude that that's the case in line with our usual standards. And now in the last just day or two, it's not just that we've also seen Tulsi Gabbard take a more extreme step, releasing a report basically suggesting the United States was involved in biological weapons labs in the Ukraine and other places. This is a direct borrow from kind of Russian rhetoric back from when they started this war early on. It's been pretty quickly debunked by people including Laura Loomer of all people, pointing and suggesting that in fact, this is really channeling some direct talking points, Russian intelligence, intelligence, and using open source information. And by the way, a lot of these are labs to do things like watching for agricultural diseases, passing through agriculture, other sorts of disease, things that, like pandemic monitoring and management that are part of what you kind of want to do to prevent something else like another global pandemic has happened in Covid. So, Julie, I want to come to you on this, obviously, as our resident intelligence community veteran. Let's start with the Havana Syndrome. That was what we originally came here to talk about. Then we have this lapse issue kind of popped up since we originally set the agenda. Like, talk to us about where these fit into Gabbard's kind of. Well, first where they fit on their merits and then how they fit into Gabbard's kind of legacy in terms of the impact she's had on the intelligence community in her time in this office.
Julia Curley
Yeah. So first I want to reiterate my disclaimer that I do not speak for the United States government on this issue. And I actually never worked on this issue at CIA. So my views on this are entirely based on the open source, unclassified record. So I think the first thing that we should talk about is that this is really a tragic human story. US personnel, including fellow CIA officers and their families, experienced something horrifying that has resisted satisfying explanation for about a decade across the US Government. I think that the total number of cases was 1500 reported. Many of the people who were involved were Russia focused officers. And many of the victim community have been publicly unhappy with the findings that the intelligence community put out in 2023 and 25 that a foreign actor was very unlikely to have caused the incidents. And I will just say that from knowing people who have written on the subject, who are some of the best analysts that I've met in the, in the business, very few judgments have caused more anguish than the AHI issue. And we've spent nearly a decade of looking for answers on this that have all come up short. Medical imaging has not found a consistent biomarker. A consistent finding has been that no single mechanism can account for all of the reported symptoms that people have experienced. But what makes this particularly agonizing is that it's possible that when this larger set of naturally occurring environmentally caused incidents, there could be a subset of foreign attacks that is causing permanent harm that could be hiding within that broader population of people who have been affected. And so this has been an agonizing problem for the intelligence community for the beginning. And it is particularly important that the analysis be allowed to proceed without politicization, which is the next thing we should talk about.
Scott R. Anderson
Yeah. So, you know, it is this difficult issue, I think, where you are in a situation where when you are accusing a foreign government of doing something like attacking your personnel, which can be taken as an act of war. In other contexts, I would say international law, that's debatable. But generally I would say there's plenty of countries going to see as begging and demanding an extremely strong response. It's something you wouldn't undertake lightly. You would set a high bar for reaching that conclusion. And that's always been part of my operation assumption reading this issue is when you look back at the binary assessments, it was that look to reach this conclusion, we got to be pretty damn sure. And while there's a lot of problematic information here, we don't reach that sort of threshold. So what does it mean when you see that rescinded along with some of the rhetoric that Gabs Gabbard has made over the years, and I think reiterated in recent days about criticizing the Biden administration, suggesting that there was some sort of COVID up here? Where does that fit in this politicization element and how does it align with these other steps that Gabbard's taken as she kind of exited the intelligence community?
Julia Curley
Yeah, so first I want to weigh the rescission on its merits should be taken seriously in directing the NIC to rescind its products. The DNI outlined five alleged shortfalls with the prior assessments. It said that there was a selective exclusion of contrary information. It mischaracterized the underlying raw reporting to suppress alternative conclusions and omitted information on source quality. This was all in a memo that had been sent to Congress earlier this week. The first three critiques go to the heart of what intelligence analysis tradecraft is. And if there were specific evidence behind them, there would be a serious problem here. But the public rescission that was issued, at least that we can see, offers no evidence of why the prior assessments fell short against those standards. Earlier this year, the DNI actually released a new Intelligence Community Directive that laid out a new set of analytic standards. But it's called ICD203. But while it's unclassified, it has not been made public. And so they're judging the prior assessments against a document that we can't see. The final two charges the DNI issued claimed that the assessments relied on on ethically flawed medical studies and on limited collection that protected the existing analytic line, that the absence of evidence ruled out of foreign involvement. These charges also may be valid or not, but the evidence behind them was not presented and could be presumably classified. Finally, I think that an important part of this is that after faulting these prior assessments without providing evidence for ruling out the Foreign act, the recall also fails to name one in this case. It doesn't contend with a potentially obvious culprit who has the motive, the placement, the access and possibly the technology to attack our officers. And so I guess I would summarize this all as this was a recall, not a replacement. The DNI didn't show her work. She judged the prior assessments by a standard that no one can actually look at. And finally, she got the benefit of denouncing the Biden administration with a without presenting our own conclusions of what actually happened. And so that's why this is particularly difficult, because the victims of this phenomena deserve a relentless search for the truth and not to be dragged into another partisan score settling by a DNI who has now repeatedly politicized intelligence. And I think that that's how we should see this rescission.
Tyler McBrien
Yeah, that's really helpful. This rescission coupled with, of course, the DNI Gabbard's past statements and actions, coupled with this Biolab story, which I would characterize as she was sort of making hay over something that's already been out in the public and was a legal program. So I definitely see the politicization motivation here and the political points that she would win among her base and the Republicans. But as many, many, many people have pointed out, this also many of her actions and narratives just perfectly track with Russian information operations. And I have to say there are a lot of accusations that she's a Russian asset. But if that's true, she has to be the worst Russian asset of all time because everyone is suspecting her to be. So I'm curious what people make of the person nominated to succeed her and whether we will see this continued dual track of politicized intelligence and actions that seem to at the very least be things that the Russians would not object to. If anyone has any, I don't know, predictions of whether this nominee will pass Jay Clayton and how he will compare to his predecessor.
Julia Curley
I don't claim to know anything about Jay Clayton. I know that it looked like he was on the fast track for being confirmed just based on the alternative of Bill Pulte who had been put forward. This is the person who would use mortgage fraud investigations to go after the President's political enemies. So a serious person with a serious background who does not show much evidence that he's going to be used as a weapon to go after political rivals would seem to be an improvement. I want to step back just to put a capstone on Tulsi's tenure as dni. And this does set a pretty low bar that it would be easy to step over for the follow on dni. Probably the latest example of it is where the DNI has cloaked what is obviously an effort to bring down assessments that the administration disagrees with in the language of integrity, of counter corruption and attempting to root out politicization and weaponization while doing precisely that to the intelligence product. This reminds me a lot of President Xi Jinping launching an anti corruption campaign that just happens to only target his own potential rivals. So we've seen this repeatedly in the IC that the DNI is overseen. We saw this with Ratcliffe with the rescission of the 2016 election assessment, recalling 16 CIA products going back a decade. And now we're seeing it with the DNI going after a product that will win political points, points with the President's base. And it's quite a fitting end to Tulsi's tenure as probably the worst DNI in the office's 20 year history and has led many people to question whether the DNI position should be abolished or fundamentally reformed given the danger that the DNI seems to pose to the nonpartisan nature of the intelligence community.
Scott R. Anderson
Well, folks, folks, that is unfortunately all the time we have today for this set of topics. But this would not be rational security if we did not leave you with some object lessons to ponder over in the week to come until we are back in your podcatcher. Tyler, what did you bring for us this week?
Tyler McBrien
I'm back on a podcast recommending a podcast.
Scott R. Anderson
God damn it, Tyler, if this is one of your other podcasts, I'm going to be particularly furious.
Tyler McBrien
No, no. I will say it may be odd to hear, but I've been in a bit of a podcast rut in terms of narrative series. But last week I was at Tribeca Audio, which is sort of the audio arm of the film festival, and was just pretty blown away by the slate. So I encourage everyone to go look at the selections, both the showcase and the official selections of Tribeca Audio. I'll just shout out one in particular that I just started. It's phenomenal. It's called A Whole Other Country. It's hosted by a reporter named Zone Zoe Kurland. It's from Marfa Public Radio. It's about this guy in Texas in the 90s who wanted to essentially secede and create or reconstitute the Republic of Texas. It's unfortunately very relevant today. I was just listening to the news about all of these referenda in Illinois to break away from Chicago. So highly recommend A Whole Other country and all the podcasts that were on display at Tribeca Audio.
Scott R. Anderson
One wonderful, great set of suggestions. I'll have to check those out as we enter the summer season of travel with lots more podcast opportunities and usually present themselves. Alan, what do you have for us this week?
Alan Rosenstein
I have barbecue technology.
Scott R. Anderson
It's stealing my lane. I like it.
Alan Rosenstein
I have as my main grill the master built Gravity Charcoal series. It's basically. It's like a pellet smoker, but instead of pellets, it's actual charcoal. It's just fence fantastic. And having owned it now for a couple of years, I decided to upgrade the temperature controller to the Fireboard brand, which is this like aftermarket, much better upgrade. And it's amazing. Like the. I AM Such a 40 year old geek dad because now I can sit in my, in my house on my phone and control my, my pit with actual good wi fi.
Scott R. Anderson
You don't even have to go outside.
Alan Rosenstein
I don't have to go outside. It's so, it's so. It's like the, the temperature. Temperatures are so incredibly stable. It's just, it's really improving my life. It's really.
Scott R. Anderson
I love it. I love it. This is my object lesson. Last week, if you may not have heard, was my fancy new outdoor gas fired pizza oven. So we're in a similar vein.
Alan Rosenstein
Which one did you get?
Scott R. Anderson
That technology I went to with the Gosney Arc xl.
Alan Rosenstein
Okay.
Scott R. Anderson
I highly recommend the Dome. I thought about it's going to be wood fired too. But I was like, what am I? I'm not going to use that, but maybe one day I'll upgrade to the Dome. But the Arc, I gotta say, it's pretty amazing.
Alan Rosenstein
I'm not, you know, for like 90 second Neapolitan cooks and even like three 4 minute New York style cooks. I'm not sure the wood does anything right. Like that's not. You're just not.
Scott R. Anderson
If you're going to bake bread, maybe maybe get a little kickery vibe.
Alan Rosenstein
Yeah.
Scott R. Anderson
But who knows. Or smoked me, which I don't know.
Alan Rosenstein
Invite me over and feed me pizza.
Scott R. Anderson
We will. Hey, we may do that. We're back in the pizza having game. We have to have another law fair cookout in my house sometime. But not today. Today I'm going to bring you a very boring object lesson. Because I am in fact somebody whose job title is in some circles, scholar, meaning I'm a really boring person who has really boring habits. And I got really titillated by something that is really boring that I'm going to share with all of you, but maybe some of you will appreciate it, and that is that there was kind of a saucy fight on the Supreme Court last week in a case that nobody paid any attention to. Which is fun if you are somebody who's interested in statutory interpretation, which is a topic that I am interested in and read it up on on occasion. And this is the case. FS Credit Opportunities Corps v. Saba Capital Master Fund Ltd. Yes. A title that lights the heart on fire with the specificity of these corporate entities suing each other over the investment company and act. But it got into a really interesting debate between Justice Barrett, who wrote the 6:3 majority, and Justice Jackson, who wrote kind of the lead dissent, or at least the part of the dissent that's, that's, that's mostly directly responding to Barrett in this sort of exchange is only joined by Justice Sotomayor. Justice Kagan kind of stepped aside and said, ah, I kind of buy it maybe, but not so much. I'm not going to lean on it here and enjoying that part of the opinion. But it's all about the use of legislative history and how Justice Jackson argues that we should be able to use legislative history and lean on this particular case to reach a contrary conclusion of which he argues is a fairly ambiguous statute. Barrett says no, there's a lot of going back and forth and it kind of gets at that question of well, if you're a textualist and you don't like statutory interpretation based on legislative history in the statutory context, then how is it and why is it you can use all of this material in terms of constitutional interpretation and that we can take the definition of what is a case or a controversy and look at huge swaths of historical tradition and practice and pretend like that is less prone to cherry picking than a defined legislative reference record? I think a fairly compelling critique that I don't see get brought to the fore that often anymore that Justice Jackson at least nods at in this dissent. So if you are interested in tax story interpretation, I think it's great. I think it's a great little teaching tool honestly if you're somebody teaching legislation or maybe even con law or a couple other topics that this might be relevant to. So worth checking out. It's just one, it's a relatively short opinion overall I think of 40 pages and this is honestly just like a three part section three, four paragraph section of or page section of Justice Barrett's opinion and I think about the same Justice Jackson's dissent, maybe even a little bit less so. Worth checking out. A nods to the advisory opinion podcast that tipped me off to this particularly interesting read as they dug into it I think last week on one of their episodes. So worth a read if you are of that particular type of nerd. And with that Julia, I've really thrown you a softball. I've given you the opposite of a dog act to follow. So go ahead, give us your inaugural object lesson. What do you have to share with us this week?
Julia Curley
Well, I've been waiting for this for almost 10 years. So let me, let me give it to you. Back in 2017-2019, I was a PDB briefer working in the Trump White House in the middle of the Russia investigation. And every day after a good 12 hour days, I would come home exhausted. I would curl up on my couch and I would bring up my mug and I would listen to rational security talk about the days affairs and it was like the deep state therapy hour that I really needed. And so because it's pride month and I'm proud of myself. I'm proud of my people in the CIA. I'm proud of the intelligence community, I'm proud of everyone who speaks truth to power. I just wanted to show you my mug and tell you how grateful I am now to get the chance to work at Lawfare as a public service fellowship and talk about the institutions that I gave my adult life to. And it's really been a pleasure to get to meet all of you.
Scott R. Anderson
We are incredibly excited to have you. For those who cannot see the video, that is a mug that has in the Ohio State certain model. A real twist of the knife to you. J.D. vance I suppose it is the Deep State logo which I want one of these mugs wherever they may be cure they're about to get a run on their whatever their sales may be. But regardless, we are thrilled to have you Julia with us now here at Lawfare and excited to have you here on Rational Security for this episode and no doubt some more episodes to come. Come until then though, we are out of time. Remember that Rational Security is a production of Lawfare, so be sure to visit lawfaremedia.org for our show page for links to past episodes, for written work and the written work of other Lawfare contributors, and for information on Lawfare's other phenomenal podcast series. While you're at it, be sure to follow Lawfare on social media wherever you socialize your media. Be sure to leave a rating or review wherever you might be listening and sign up to become a material supporter of Lawfare on Patreon for an ad free version of this podcast among other special benefits. For more information, visit lawfordmedia.org support our audio engineer and producer this week was Noam Osband of Goat Rodeo and our music as always was performed by Sophia Yan and we were once again edited by the wonderful Jen Patcha. On behalf of my guests Julia, Tyler, Ari and Alan, I am Scott R. Andersen and we will talk to you next week. Until then, goodbye.
Michael's Party Shop Announcer
Is your kid's birthday coming up? Don't stress the prep. The Party Shop at Michael's is your one stop shop for everything from Bluey to Rodeo. Transform your space into a birthday wonderland with dreamworthy tablescapes and decor starting at $0.99 plus get free helium inflation on select balloon styles. Whether you're crafting a one of a kind bash or grabbing the essentials, Michaels has everything to make their big day legendary. Shop now and order ahead for delivery or in store pickup. Michaels everything to celebrate anything.
Date: June 18, 2026
Host: Scott R. Anderson (Lawfare Institute)
Guests: Ari Tapatabai (contributing editor, former public service fellow), Julia Curley (Public Service Fellow, former intelligence community/NSC official), Tyler McBrien, Alan Rosenstein
This week’s edition of "Rational Security" brings together the Lawfare team to decode a trio of urgent U.S. national security stories. The episode’s central theme, reflected in the “Predestination” title, is the frustrating sense of inevitability—from the U.S.-Iran war’s anticlimactic peace deal, to government overreach in AI regulation, and the politicization of intelligence. The panel brings both substance and signature Lawfare irreverence to bear in dissecting the week’s events.
[06:00–38:14]
Panel: Scott R. Anderson (host), Ari Tapatabai, Julia Curley, Tyler McBrien, Alan Rosenstein
“Did a lawyer approve of this? …I do not think that our lawyers would have let us…leave the room alive if we had walked away with this kind of text.” (09:50)
"Did we lose the war in Iran? It really feels like we just lost the war in Iran." (14:14)
"The United States won every battle… but the United States lost this war. We lost 14 servicemen, we depleted critical munitions, we alienated our allies…This has been a strategic disaster." (19:05)
“…we are now negotiating with and giving billions to [a regime that] is using the war as a cover…to further continue to repress people…” (15:51)
"The administration…sent a message to the intelligence community that dissent would be punished...Analysis at every phase of this was shut out, punished, or ignored. So this is the price of politicized intelligence…" (32:02)
“…this seems bad for the Jews...there does seem to be at least a kernel of truth to the idea that America got into this war in part because Israel pushed it. Which…doesn’t justify antisemitism. But…it’s not a great data point.” (33:04)
[44:51–64:25]
Panel: Alan Rosenstein, Scott R. Anderson, Tyler McBrien, Julia Curley
"This is a very, very big deal… the beginning of essentially a de facto licensing regime for frontier models.” (46:45)
"It's notable to me that the letter does not appear to do what the Commerce Department is saying it does...[This is] a basic competence question." (51:55)
“If in fact Fable is the sort of model that triggers emergency national security freakout... every single model from here on to forever is going to be considered this dangerous.” (58:00)
[64:25–75:45]
Panel: Scott R. Anderson, Julia Curley, Tyler McBrien
“Very few judgments have caused more anguish than the AHI issue...the victims...deserve a relentless search for the truth and not to be dragged into another partisan score settling...” (70:00)
“This was a recall, not a replacement. The DNI didn't show her work...she got the benefit of denouncing the Biden administration without presenting [her] own conclusions.” (70:00)
"Probably the worst DNI in the office's 20 year history and has led many people to question whether the DNI position should be abolished or fundamentally reformed given the danger that the DNI seems to pose to the nonpartisan nature of the intelligence community." (75:45)
“We don't know how the hell you got our phone number, but we'll take it.” (05:29)
“Because it's Pride month and I'm proud of myself, I'm proud of my people in the CIA, I'm proud of the intelligence community, I'm proud of everyone who speaks truth to power…” (81:19)
This “Predestination” episode finds the U.S. in a state of anxious drift, with leadership challenged at home and abroad. The Lawfare team’s unsparing analysis cuts through government spin and wishful thinking, whether on Iran, AI, or intelligence integrity—leaving listeners not with reassurance, but with urgent questions about what real security looks like.
For deeper dives, referenced Lawfare articles, and transcripts, visit lawfaremedia.org.