Silicon Valley’s secret industrial spy war

Danny Crichton:

Zach, thank you so much for joining us today.

Zach Dorfman:

Pleasure to be here, Danny. Thanks so much.

Danny Crichton:

So, you recently wrote a piece on political magazine called Moscow's Spies Were Stealing US Tech-Until the FBI Started a Sabotage Campaign, and this went pretty viral all across I think Silicon Valley. A bunch of friends of mine shared it with me. Why don't we just dive into how did you find the story, how did you run into it, and what's the story about?

Zach Dorfman:

How I found the story is that I'm based in San Francisco, and I'm a national security reporter, which makes me like an odd duck because-

Danny Crichton:

You're the only one there. I mean-

Zach Dorfman:

Essentially, yeah. Essentially. There are some other very talented journalists who do cybersecurity reporting, and of course that has overlap with the world of espionage, but I'm just like an odd generalist national security reporter in the Bay Area. And years ago, when I moved from New York to San Francisco, I had the, I don't know, the intuition, I guess, that there was not enough reporting on Silicon Valley from just a straight national security perspective. And I was in a great position, because there just weren't many other reporters out there that were really trying to dig on that stuff. And so, what I did was I did what reporters do and just started reaching out to retired intelligence community people, who had either dealt with Silicon Valley and the Bay Area, or had been based for long periods of time in the Bay Area. And I met a group of folks who were longtime FBI counterintelligence agents who had spent, sometimes their entire careers essentially, in San Francisco and Silicon Valley.

These are folks that spent 20 or 30 years only doing counterintelligence work in the valley and in the Bay Area, and I got lucky because they had some stories to tell. Honestly, some of them had been waiting their entire careers for somebody ... It would be like to come by and say, "This is really interesting and important, and clearly there's a lot of foreign spying in the Bay Area and in the valley. Can you tell me about some of it?" And luckily, some of those folks had been retired for a while, and they felt comfortable really leaning in and telling some of the stories from the Cold War, because enough time had passed, the sensitivities weren't quite there the same way. And one of those FBI agents, by the name of Rick Smith, became somebody that I-

Danny Crichton:

Is that his real name?

Zach Dorfman:

That is his real name. Rick Smith is his real name, he's a great guy. He still lives in the Bay Area, he lives in San Francisco. He spent decades chasing Soviet and then Russian spies around San Francisco. He had this amazing story to tell, and this story was that basically in the early 1980s he ran a massive sabotage campaign, where he personally recruited a Silicon Valley entrepreneur, who was Austrian but lived in the valley, to run this high-tech sabotage operation where this Austrian, this entrepreneur, would sell microchips and other microchip production equipment to the Soviet Bloc, but a great deal of it would be tampered with. It ran for years, they ended up selling millions of dollars worth of tampered high-tech to the Soviets, and it's an amazing story and I was just fortunate enough to have Rick feel like he could talk to me about it.

Danny Crichton:

When I think of the context here, so in the early '80s, Soviet Union and the U.S. actually had a little bit of trade. So I mean, the Iron Curtain is up, but it's not like U.S., China where there's hundreds of billions and the top two trading partners for both countries. There's a couple billion, so a good chunk, and a chunk, a piece of that was, I think it was about $200 million of microelectronics that were coming from Silicon Valley into the Soviet Bloc. And you had a interview with Chris Miller from Chip Wars, who's been on the Risk Gaming podcast, he's a former co-author of mine, and Soviet historian, Soviet background. And the Soviets really struggled this category around computer science and building up an industry here, so they needed these parts and needed them desperately.

Zach Dorfman:

They could not indigenously produce what the valley was creating, and what was revolutionizing a microelectronics industry at the time. Like you said, there was I think a couple hundred million dollars of high-tech trade, but again, it was a fraction of a fraction of what you get now between the U.S. and China. Nonetheless, the Soviets understood very well, even at the dawn of the 1980s, that there would be significant military and intelligence advantages to the countries that could indigenously produce these microelectronics, which were being put into missiles, radar systems, infrastructure for space exploration. Obviously processing power was rudimentary then compared to what we have now, but was then cutting edge, and they realized they were quite behind. So, they had a two-prong strategy. One was they were going to steal as much of it as they could, but they also wanted to reverse-engineer it and create their own factories that would somehow produce these produce microchips and other equipment from these stolen designs.

So, what this operation that Rick Smith ran did was it didn't just sell them computer chips that had been tampered with, it actually sold them highly specialized computer production equipment that they were trying to purchase surreptitiously in order to open up their own computer factories. This was all huge failure, and at least part of the reason for that failure was this campaign, this FBI campaign. And I think it's fair to assume other associated campaigns that were going on at the time, because the FBI and the CIA were more than aware of what the Soviets and their allies were trying to get their hands on.

Danny Crichton:

Let's dive in. So there's this operation, which was called Operation Intering, I don't think you describe where that name comes from, or maybe you didn't discover. Where did the name come from?

Zach Dorfman:

My understanding is it was just short for international ring. Not the most creative.

Danny Crichton:

Yes, feels very FBI, Project Alpha. Yeah.

Zach Dorfman:

Exactly. This is the early days of this stuff, so this was really the first historical period when they were really engaged with this. The late '70s and early '80s was when the high-tech war and these massive covert, illegal importation schemes through deniable parties, and third parties in third and fourth countries and all that stuff, that's when that really came to the fore. So I don't know, I think it was probably just very novel at the time.

Danny Crichton:

I'm totally doing this on the spot. I mean, I think of it as, one, you had the rise of electronics in the '70s, so for the first time there was these inventions that other countries didn't have and they needed. And so, that created a dependency that obviously was a asset to use if you could take advantage of it. Two, this is post the church commission, which they believe is '75, '77?

Zach Dorfman:

Yep.

Danny Crichton:

Cleaning up the intelligence community, moving it away from domestic surveillance and the 1960s counter Vietnam. I just wonder if some of those resources that were so focused on American citizens, and fighting all the anti-Vietnam protesters and civil rights activists, that some of those resources got redeployed. And so, suddenly you had a little bit more capacity in that era, two, three years later after the rules were changed.

Zach Dorfman:

I mean, I can't speak to that directly in terms of the redistribution of resources within the intelligence community, although I'm sure that with the sun setting of dictante, because you had a decade where you had Nixon and Kissinger, and then early Carter too followed along that, where there was this attempt to find this modus vivendi with the Soviet Union. There was always an understanding of there were adversarial interests, but you had this realpolitik that dominated U.S. foreign policy, vis-a-vis Moscow and its allies, and that really ended when the Soviets invaded Afghanistan. So, you had this long period where you had this nascent microelectronics industry, you had this foreign policy, that at a grand strategic level was more interested in finding common ground, and some of that was through mutually beneficial trade. So, there's all that stuff going on. Then meanwhile, by the end of the 1970s and early 1980s, you have more intelligence coming in from CIA, FBI sources, and also French sources.

There was one very important French source, a KGB officer who was a senior science and technology officer by the last name of Vetrov, who produced something known as the Farewell Dossier. And the Farewell Dossier showed basically that the Soviets had been on a giant high-tech theft spree from the late '70s onward. And when the U.S. got that, they also thought, "Okay, all right, there's a different game that's been played than we've been really aware of." We did not understand the full extent of the Russian attempts to steal and replicate American high-tech. So, you have all that stuff coming together in the early 1980s. And then of course the transition from Carter to Reagan too, really underlying that, where then you have the whole equal empire stuff, and it was like game on for the FBI. And you actually hear some of the old retired agents talk about it like that, where they felt like they had a lot more room to maneuver by the early 1980s than they did for years before.

Danny Crichton:

Interesting. And you emphasize this, so I mean, this operation wasn't just about sabotaging equipment as you described, but it was also collecting information on what do the Soviets want, what parts are they asking for, what technologies are they interested in? And getting sort of intel that way, sort of market demand information. And then the second piece here, as we mentioned with the building machines, was not just sabotaging the dual use, and the missiles, and wherever you wanted to put these chips, but really undermining from the beginning the indigenization of this industry, which the Soviets had really prioritized with the 1980s.

Zach Dorfman:

Right. I mean, if you are subtly altering chip production technology, you are creating this long tail, where the folks who are building those, I mean, imagine the amount of human and other sorts of capital that go into creating an entire industry. It's like the old import substitution model, I mean, but you're doing that at a massive scale and you are bringing the state's resources to bear massively on that too. It's security services, as well as it's economic might. And there's the short-term ramifications of it, which is that some chips didn't work, the chips they imported didn't work. But when you're talking about it a few levels up, you are basically handicapping, you're kneecapping entire industries way, way, way upstream. And in theory, that's what Intering was trying to do. And as far as I know, I'm not an expert on this, but I don't think the Russians really ever succeeded in an indigenous chip industry of the sort that you have in the United States.

Danny Crichton:

No, up until the present day there's not a very deep industry there. And it's interesting because, I mean, if you had an iPhone, you could turn it on and see that it's working, not working, or whatever the case may be. There's a lot of devices that work like this. In the chip industry you get this tool, and you have to use it. It's a very precise tool. And the way you described it was if you move just a couple microns off, is that just a miscalibration that you're doing? These tools require very excellently trained workers to operate them, so is it just, "Look, we haven't had the tool before. We have the tool, but is it our training that's off? Is it the electricity in the building?" All these different factors. So, the possible deniability of it not functioning is, in my view, very unique to this high precision industry that doesn't apply to a lot of other technologies that might've been also influenced by an FBI, CIA operation.

Zach Dorfman:

Well, yeah, I mean, I think that's a really good point. I mean, you want it to be potentially multi-causal or over-determined in the way that-

Danny Crichton:

Exactly.

Zach Dorfman:

Because then you get to keep the operation running. And one of the most interesting parts of this story, and reporting it out, and speaking to FBI agents about it was asking them these sorts of questions, which was like, "Okay, well if you're sabotaging the technology, they're going to know right away. How do you keep an operation going? What are the trade-offs that you have to deal with when you are running this ..." It's technically called an offensive counterintelligence scheme, because if everything doesn't work, obviously they're going to know that something's wrong immediately. But if you let too much unadulterated technology through, then is it possible that on balance you're actually providing a net benefit to your adversary? And what they ended up doing too was that they ended up sabotaging the technology in different ways to cover their own tracks. I had one FBI agent tell me, "Some of it we made it seem like it just got messed up on wear and tear." It was literally somebody shook the Amazon package too hard, you know what I mean? Like bad delivery?

Danny Crichton:

Like my courier, yes.

Zach Dorfman:

Exactly, exactly. Right, right. My courier in Geneva, or whatever. So, some of it was like it would just show up and be like, "Oh man, these are just kind of damaged." And then some of it was what you described, which was ultra subtly altering the technology itself so it would be imperceptible to the assumedly, very technically savvy people that were running those factories, or were doing that science and technology intake for the KGB and its allies. And they managed to make it work for years, until eventually and inevitably the Soviets and their allies suspicions were aroused, and then the operation had to end. But I mean, the fact that it ran for years at all, and they got millions of dollars worth of technology through is itself extraordinary.

Danny Crichton:

And one of the things that comes out of this, I mean, obviously you're sending all these technologies over there, a bunch of ethical issues come up that they debated throughout this operation, one of which you just talked about, which was what if we have too much good technology to get behind the Iron Curtain? Are we just raising the cost? If one in four of the devices works, they throw away the three bad ones, they keep the one good one, you've quadrupled the cost and that's bad for the Soviets. But if they had no access in the first place, then qualitatively you've actually helped rather than harmed their ability to move forward.

Zach Dorfman:

Precisely.

Danny Crichton:

Putting a pin real quick on that, I think it's amazing just from a media perspective of the leak that's possible that goes on here in a way that maybe was easier in the '70s, '80s to hide, and would be a little bit harder with the internet, and blogging, and independent researchers or journalists like yourself, where it's like, "Well, guess what? We're just shipping the chips ourselves. That's who's helping the Soviets out. The U.S. government, the FBI is doing it." And I thought that was super interesting, so keeping this very under wraps. But then there's a wider question around the ethics of this, which was you didn't know where the chips were going. And so, it's assumed it was going into a weapon system, or a missile, or into the Soviet army, but could go into hospitals, it go into a lot of other places? And I thought it was interesting that there was quite a lot of discussion around those sorts of ethical issues at the dawn of the computer sciences, we're trying to think about where all this technology ending up all across society.

Zach Dorfman:

So yeah, I mean, I think there's two points, and I'll address the latter one first because it's ... It was interesting to me, as I think it was ... I'm glad you brought it up, but when I raised that question to the former FBI agents who worked the case, it was interesting to hear that they had really thought about this, and that this was just something that they were quite worried about. I mean, they admitted to me, like, "We didn't know 100% where this ..." Once that stuff goes behind the Iron Curtain, we assume we know what their uses might be because we understand what their intelligence requirements are, from things like the Farewell Dossier and other things, but they didn't know for certain. But they eventually concluded that they were just buying that stuff to take it apart and reverse engineer it. But of course that's a danger.

I mean, you could absolutely imagine a scenario where some of this stuff does get rerouted for civilian uses, and then you are being put in a position of real moral hazard. And there are examples through history of these double agent operations going awry. I mean, there's two that I think are worth mentioning. One was a CIA operation that was reported by James Risen in his book State of War, where the U.S. decided it wanted to sabotage Iran's nuclear program by having a Russian scientist who actually worked for the U.S. to pretend like he was defecting, to then provide the Iranians specifications that could help speed up the nuclear program. Now, what was supposed to happen was the specs were supposed to be altered in a way to arrest the program, but the allegations in Risen's story are that in fact the specifications were not altered enough.

And that in fact, instead of arresting the program, it supercharged the program. So, you actually had the U.S. provide the Iranians information that helped them lead to a breakthrough in the nuclear program. Less known, but I think is almost more shocking, is a Cold War one by David Wise, the Cold War National Security reporter, who wrote a book that dealt with the Soviet government's chemical weapons program. And in the 1960s, the FBI ran a double agent, who was a military, American military official, who was recruited, "recruited," not really recruited, but recruited by a GRU officer in the U.S. And this U.S. military official had access to, I forget the name of the base, but whatever base I believe was in Maryland that did the U.S.'s chemical and biological weapons research. And once again, this U.S. army official, who was supposed to be working as a double agent on behalf of the U.S., provided "altered" information to the Russians.

But once again, it was not altered enough and it led the Russians, the Soviets to breaking through and developing Novichok, which is the deadliest nerve gas known to man, and has been used repeatedly by the Russians over the last few years to try and poison defectors. So, that's a long cul-de-sac from what you asked, but it's important to know that there are known real-life examples of these double agent operations going awry where the U.S. has tried to do one thing and has actually ended up doing the opposite.

Danny Crichton:

Well, I mean, what's interesting to me is there's the mechanics of espionage, counter espionage and what's going on here, and then there's this wider piece around knowledge and the sharing of knowledge. And even today when we think about science technology, and we just assume everything's open source, everything flows out, everyone knows everything, William Gibson says, "The future is here, it's just not equally distributed." But the reality is, the internet really distributes a lot of information very quickly, but that's still not the case even when you get into these machines, and these parts, think even 2024, there's so much tested knowledge that's held in just a couple of heads that keeps everything in the world operational, that even today you can still do these sorts of things. And the question becomes this game of like, well, how much do you have to adjust?

Because it's almost, you're on the Risk Gaming podcast, but you can almost imagine as a game theory is like, well, I want to make it different enough so that you can't reconstruct it or be inspired enough by it, where smart people looking at the same information goes, "Well, this feels a little bit off. What if we went back the way we just came?" And then you're suddenly like, "Eureka, we found exactly the right solution," and you led the bread crumbs to this endpoint that you started in the maze. And so, I just find that super fascinating, because to me that's a very ... Obfuscation is not a field that is studied in universities. There's no discipline for like, hey, how do we take this knowledge and make it harder for people to understand, outside of maybe some recondite professors writing their abstruse theories. But to me, it's a super fascinating game of how do you occlude that knowledge from folks who are otherwise desperately seeking it?

Zach Dorfman:

Boy, I think that is such an interesting question, and I think what makes it particularly complicated is that you're talking about the nexus between thinking like a spy, thinking in terms of multi-layered offensive counterintelligence related schemes, and then also thinking like an engineer, a technician, a nuclear scientist. And there is so rarely the intermingling of that knowledge at that level. How do you carry out grand strategic objectives through counterintelligence schemes, and then at the same time do so in a way that technically apt, but in a way that the person who you might be consulting with, who has that highly technical knowledge, can think through the game theoretic scenarios, where somebody who is their colleague on the opposite side would make exactly the wrong decisions that you want them to make without ever actually having any hints about the right ones, that you don't ever want to see them understanding the possibilities?

I don't know. I mean, I wonder if in some deep, dark corners of the U.S. government there are people thinking about this and this nexus. I have never seen anything in the public sphere related to this, because I think it's so complex and so sensitive.

Danny Crichton:

And I think it's a good example. I mean, obviously stove piping is a theme across the intelligence community going for back decades, pre-9/11, post-9/11, it doesn't matter how many reforms you do, there's too many agencies, too many people all connecting the dots. But it does seem like this is a category where you're bringing together a lot of specialization across a lot of different fields, both within the IC as well as private industry, who, as you noted, were very intrinsically important to this particular operation, I'm sure similar ones that have happened in the modern era, not just in the U.s., by the way. I mean, in your story, I believe people were in London, people were in Switzerland, there was this infrastructure both on the logistics and supply chain just to make it all work.

And so, the level of cooperation you needed, and secrecy in that group, but it's so decentralized, I think is really exceptional. And then, I just throw out that the only pop culture reference I can think of is Rogue One, the Star Wars story, which is about not just the Death Star plans, but putting in a fatal flaw by the weapons designer designing this massive battle station where everything falls together, everything's being cross checked, but there's a secret flaw that they've introduced into this that someone else could potentially exploit in the future if they were to send the torpedo in exactly the right hole on the top of the surface of this moon. And to me, it's a little bit like that. It's one of the only movies I point to is it's an engineering focused film. It's about how engineers can make the world a better place.

Zach Dorfman:

How did I never put that-

Danny Crichton:

We don't get a lot of those at all.

Zach Dorfman:

How did I never put that together? Yeah, of course. That's absolutely right. That's absolutely right. I mean, an interesting corollary to what we've been talking about a little bit, which is there's the famous Stuxnet example, but I mean, that's a virus, that's a software issue. But there are potential scenarios that you can imagine a ... And in fact, there has been some reporting about this vis-a-vis the North Korea nuclear program. The New York Times has done some reporting on this in past years, David Sanger, where there have apparently been longstanding supply chain interdiction, covert actions, where they seem to have sabotaged the supply chain in a way that would systematically cause miscalculations when the North Koreans attempt to actually do ballistic missile tests. That there's these flaws that have been introduced en masse. That probably exists at a much greater scale than we're aware, honestly.

But that's also something that states are incredibly loath to talk about. And anything that's even anywhere near current, I mean, people just button up entirely on it. You just cannot get U.S. intelligence officials to talk about it, because if it's happening at its scale, I mean, again, think about some of the moral hazards we've been talking about. If there are flaws, like well, we do know for instance that the U.S. ran a company for many years in Switzerland called Crypto AG, which created fatal flaws in the hardware of encryption technology, which is, talk about holy grail. I mean, that's incredible stuff if you're a spy agency, but there are other cases you can imagine where things get into commercial technology en masse, and then you're really talking about some difficult ethical questions.

Danny Crichton:

Well, I think, I mean, on the flip side, you have Jan Marsalek, the former COO of Wirecard, who was at the center of a DAX 40 company in Germany, a massive FinTech high-flyer, which we increasingly, increasingly are figuring out is one of the most powerful agents within the Russian intelligence operations was at the heart of Europe running ... I mean, every week it feels like there's another major blockbuster story of another operation he was at the center of, and meanwhile running the most powerful payments company in Europe. And so, imagine the place you have, the amount of data you are able to collect. To me, that crossover between commercial into the espionage community, let's call it, is just super fascinating.

Zach Dorfman:

I mean, you and I are weirdo nerds who know about that story, and I think that has been systematically under covered, because I think the idea of a payments processing giant in the heart of the European Union, the head of that company being ... The report, as far as I understand, is a little bit muddy whether he was just a recruited asset or was himself like an FSB officer, like a Russian intelligence officer. Either way, that is scandalous at a level that's hard for me to comprehend. And also very much, it's a warning, but it's also extraordinarily illustrative about the value that intelligence agencies place on high level recruitments in the tech world, and the value of having somebody in a position like that, like at Wirecard.

I mean, they were moving money surreptitiously for the Russian intelligence services all over Europe. I mean, it's incredible what you can do. I mean, you just have to be a little creative about who you think a great intelligence officer would be. Like yeah, you want somebody in the State Department? Sure. Yeah, you want somebody in the treasury, but if you want to understand the modern world, you have to understand the tech world. And to do that, if you can actually have seeded penetrations of major tech companies, I mean, you're doing great. I mean, look, I have no love for the Russian intelligence services, but purely from a perspective of a valuable recruitment. I mean, they got game.

Danny Crichton:

Well, and we just saw recently with the hostage exchange, recording this was about a week ago, with Paul Whalen and Evan Gershkovich from the Wall Street Journal getting sent here. And then in exchange, we sent prisoners back to Russia or released them from Europe. And one of those were afraid of the Americans, very deep cover family where the children didn't even know what was going on. They show up off the plane, arrive in front of Vladimir Putin. Putin actually speaks Spanish to them because they've never spoken any other language, they had no idea what the concept of why the Russian president was showing up there. And so, I do think it is interesting to see the full scale and full range of this. And then, towards your point about tech asset and recruitment, I mean, we just saw a former Twitter employee who was found guilty of working with the Kingdom of Saudi Arabia.

We see similar things in the telco industry. And so, I mean, as a reminder, because you talked at the very beginning of the show about how there's not a lot of national security reporters, mostly cybersecurity who cover Silicon Valley, this is a reminder of what we always talk about with insider threats. It's not just about other companies stealing information, but it is like someone who has access to your entire database, if you're a payments processor, a health provider, I mean, that can be useful to all kinds of people around the world who may be willing to pay, sometimes an obscene amount of money in order to acquire that information.

Zach Dorfman:

Oh, of course. I once had a former CIA officer say to me, "Look, in the world of espionage it's already the 22nd century." He was like, "I don't know what you guys think is going on, but what we're doing?" And that stuck with me. And look, I don't think it's any surprise to anybody who's a close watcher of the valley, that the collation of big data sets in the commercial context has become increasingly important. And in the intelligence context, it has become overwhelmingly so. There's a way in which what is valuable to an intelligence agency on the commercial side is viewed in a siloed context, but that I think is very, very short-sighted. And so, folks who are working in different areas, med tech, FinTech, they might not see the intelligence uses of their product, but that doesn't mean that intelligence services don't see the intelligence value of their product.

I think insider threat gets attention, but at the same time I still get the sense, and I'm not a reporter who reports on the companies as much as such. In fact, I'd like to do a little bit more from the insider thread side of things, because I think sometimes those companies see lots of things that they don't actually end up telling the FBI. Maybe they don't realize what they see, but sometimes they see it and they don't want to deal with it for a variety of reasons, which has created a lot of tension over the years with the bureau and the valley. But I think that the scope of potential interest is much wider than many of those folks even realize. And in the years to come, it'll be really fascinating to see, because folks are going to get popped.

There's going to be hacks, there's going to be human penetrations of tech companies, and it's going to be really interesting to see which ones actually become the focus, because it'll be like a leading indicator of where the spy services are going. And that will be, I don't know, that'll be interesting on its own there.

Danny Crichton:

Well, I mean, obviously there's been a lot of, I would call tension. The FBI has done a lot of educational missions to Silicon Valley. There's folks who are trying to do that full-time, but to larger big tech companies, smaller startups, I would say the vast majority of people are relatively willing to hear the story and connect. There are definitely companies who won't for a variety of reasons, who don't want to get tied up with this or frankly have a lot of business interest in sensitive countries and so, they don't even want to be near it. Ironically, they probably talk to the security services in those countries, which is a little bit more mandatory than our more voluntary system here.

But I mean, they've certainly been trying, but as always, when you're building these companies, you have that data set. Let's say you're building an AI company, you have the data set, but spending money on securing these systems, securing the data is expensive and doesn't drive revenue. And I mean, that's where I think that tension is for a lot of Silicon Valley companies on this specific issue is you can be very sensitive, it's very valuable to a foreign adversary, but there's almost no incentive in the system for you to fix it.

Zach Dorfman:

Yeah, that's 100% right. I mean, I think it's also you can view a certain amount of IP theft as the cost of doing business, and do you want to jeopardize your work in another country by making public or raising a stink about the fact that that country's intelligence services might've stolen something from your company that might affect your ... And in fact, it might actually be worse for your IPO if you go public with the fact that they've stolen it than just letting that be-

Danny Crichton:

Well, I would even put it more bluntly, which is most companies, I mean, IP protections are de minimis. I mean, you do what you can, but the idea is you got to iterate fast, you just got to keep ahead of the competition, you move faster than anyone else, you move automatically-

Zach Dorfman:

Right, because they've built something four models ago, so who cares, right?

Danny Crichton:

... and particularly in exponential spaces like AI, we'll see how long it stays exponential, but if you can just stay a little bit faster than everyone else you just get advantage more, and more, and more. But I think it's interesting because I mean, in the Lux portfolio a lot of our companies are in defense, they have ITAR requirements, we go through a lot of training for those companies to make sure they follow all Defense Department guidelines. Everything that gets exported has to go through controls. So I mean, the challenge in defense is often the opposite, which is you can't export anything to anywhere without bureaucracy. And at the Pentagon, which generally means that our companies are the slowest compared to, I'm thinking of Turkey or Korea, who have much less controls, which are able to export to a variety of other countries. And so, there is that sliding scale of how do you create companies that are nimble and are able to grow, certainly to our allies, but aren't just moving it all the way to willy-nilly like everyone has all the technology in the world?

Zach Dorfman:

I empathize with companies that are in ITAR controlled space for that reason, because there are reasonable critiques on the Silicon Valley side about what working with and through government does. Newsflash, the Pentagon moves like molasses. There's a reason why there've been attempts over the years, like DIU, or was it DIU or DIUX? I forget that they've gone, but there have been attempts over the years to create mechanisms to speed those things up. I haven't done a ton of reporting on this, so I would hesitate to say that it's been a failure, but I think it's been complicating. Yeah, I see what you-

Danny Crichton:

No, I mean, we're on the receiving end. It has not gotten better. It's nice, and they're all good people, and Raj Shah just published his book and his experience of standing up DIUX, the joint artificial intelligence center, we had first founding director on those Risk Gaming podcasts a little while back. There are attempts to try to make these faster, but you're starting from a perspective of no leaks, no security problems. Actually, it's really hard to grow anything. I mean, the best way to not have a train crash into a car is just not run a train or a car, then you can't have a crash. It doesn't exist. And in the real world, we have to actually build stuff to be able to solve problems. And so, that's where the tension source starts.

Zach Dorfman:

And the Pentagon does not operate by market incentives. Right? I mean-

Danny Crichton:

It does also operate with a very good risk perspective. I mean, that's the challenge is what is really the risk here? But understandable. But I know we only have a couple minutes, so I want to pivot the conversation to some of the previous work you did, because you and I first started having an online dialogue a little bit around your flagship reporting around China, specifically 2012, and the CIA's massive intelligence gathering network that was on the mainland, and then this roll-up that the MSS, China's internal security agency, Ministry of State Security, was able to figure out how the CIA was connecting with its operatives locally in China, and very rapidly identified all of this. And it happened right around the time that Xi Jinping was transitioning into power. It doesn't happen instantly in the Chinese system for him, but got one role, second role, and took over the whole system.

And I'm curious because it's been a couple of years, let's not summarize it, but you had this groundbreaking piece a couple of years ago. I'm curious if there's been any echoes, reflections a couple of years down the line now since those pieces came out, about the implications there. Because when I read this, I felt that this was definitive and arguably one of the determinative factors in why Xi Jinping went the route that he did over the last 10, 12 years since coming to power, is that he comes to power and realizes there's this massive CIA operation all around him, up to potentially people near him, and that colored his perception of U.S. China relations straight going in. But I'm curious if you've seen any follow-ons, either from other journalists or people you've talked to about some of the work that you did on that story?

Zach Dorfman:

I can say without question, there was a sense within the U.S. intelligence community that the revelations by the MSS about the degree and the depth of U.S. CIA penetration of the Chinese government helped drive Xi into a more authoritarian space, not just in terms of a adversarial relationship with the U.S., but also many of the purges that greeted the first few years in particular of Xi's rule. There was a sense that the systemic corruption in the Chinese system had created the space for infiltration by foreign intelligence services, the CIA foremost among them. The anti-corruption drive was, I think misdescribed a lot. I mean, where the deep causes of it were not remarked upon, I think sufficiently enough. Because I mean, it was good rhetorically for Xi to come out and say, "Look, I know that there's petty criminality everywhere, and these folks in the local party cadres are not working for you."

But higher levels, a lot of this was driven by this sense that corruption had led to foreign penetration, and therefore the increasing authoritarianism was, at least partially, a direct cause of this explosive revelation within the Chinese system that was a huge shock to it.

Danny Crichton:

I mean, I would love to see a book on this. I don't think you ever wrote it up as a book.

Zach Dorfman:

I tried, but I couldn't get any takers. I was like, "I think this is worthy of a book," but yeah.

Danny Crichton:

Now you're just pissing me off. But to me, it's not on a footnote, it's a full chapter of the timing of that could not have been worse in my view from the U.S. If you did this five years in you could have said, "Okay, I trusted folks and now I understand," but it almost happens contemporaneously. It's almost, not literally down to the day, but at the exact same time you're coming to power realizing that everyone around you might actually be a foreign adversary, is actually getting funded in a secret bank account, what does that do to your psychology, to your views on trust, not just overseas but to people in your own office?

And I've seen no one capture that story as a turning point in his life. I mean, Xi, before becoming president was vice president, had visited the U.S., and there was very little sign of what was to come. And I've always been curious, I mean, this is always speculation, but I've always been curious of whether this was a trigger and we're saying, "Well, he had all these dark thoughts and just never expressed them at the time," as opposed to saying, "Well, he changed his mind, and he collected new pieces of information that came right as he was taking over in power that radically changed the course of history."

Zach Dorfman:

So, I'm actually really glad that you brought this up because I haven't talked about this in a long time, and I think you have identified the geopolitical import of that store beyond it being like an apocryl intelligence failure on the U.S.'s part. There's two things. I mean, one, look, it's the CIA is job to recruit foreign sources and foreign governments. And in terms of holy grail, you want to recruit high level sources in China and Russia. I mean, obviously. That's what you do, that's what your job is. They were doing a good job. The technical aspects of the covert communication systems they use were deeply problematic, and that led to people being rolled up, and killed, and imprisoned. But that said, the fact that they were knocking on an open door to recruit Chinese government officials, that's their job.

But something you said made me recall a conversation that I had with a senior CIA official, and I did a three part follow-up to the original foreign policy and Yahoo news pieces. I did a three part follow up in foreign policy that went through, that looked at some of this stuff that you described. And what this official said to me was, I asked him about this, like, "Well, did you ever think about the potential blowback?" And this person said to me, "We did. And we also thought about what would happen in the U.S. system if an American president discovered that there were dozens of high level Chinese penetrations of the U.S. government." How would we react? And then, now obviously the two systems are quite different, and you don't want to overly mirror image. Nonetheless, these are people we're talking about. Systems do react to major exogenous shocks, like finding out that you are comprehensively penetrated by an adversarial spy service.

So, I think it is worth thinking through that, one, how the United States might've reacted, and two, in the larger context of who do you recruit, and how, and how you maintain that secrecy and the safety of those assets, not just for the safety of the sources themselves, which is at paramount, but also to prevent that kind of blowback that you're describing. Because then you have to weigh, was that secret intelligence that we obtained worth the generational shift in superpower relations that resulted when it all blew up? And these are very difficult questions to answer, but I think they're ones that are worth asking, and they almost never get adjudicated-

Danny Crichton:

Never get asked.

Zach Dorfman:

... in the public sphere, ever.

Danny Crichton:

Yes. Well, thank God you're doing this work. Thank God there's someone out there who's connecting the dots on Silicon Valley technology and geopolitics. But Zach Dorfman, thank you so much for joining us.

Zach Dorfman:

Danny, thank you so much.

‍