To watch the video, click here.


Peter Robinson: A college senior has one job offer from Google and another from the CIA. Which job should she take? We'll come to that. But first, an examination of the entire vast American intelligence apparatus. Amy Zegart, on Uncommon Knowledge, now. Welcome to Uncommon Knowledge. I'm Peter Robinson. A fellow at the Hoover Institution and a professor of political science at Stanford, Amy Zegart served on the National Security Council for President Bill Clinton, and advised the 2000 Presidential Campaign of then governor, later President George W. Bush. Amy Zegart's new book, Spies, Lies, and Algorithms, the History and Future of American Intelligence. Amy, welcome.

Amy Zegart: Peter, thanks so much for having me.

Peter Robinson: First question, let me set it up with two quotations. Quotation number one, Michael Rubin of the American Enterprise Institute writing about the withdrawal from Afghanistan, which he calls an intelligence failure. Quote, not only did the Central Intelligence Agency and other U.S. intelligence agencies wildly underestimate the speed of the Taliban advance. They appeared to have been blind to the political dealings of the Taliban and the military prepositioning the Taliban had achieved, close quote. Here's quotation two, Julian Barnes and David Sanger in the New York Times, the United States intelligence agencies, unearthed Russia's war plans. They accurately assessed Putin's intentions. They got the timing of his invasion right, almost to the hour. The success of American intelligence, the success of American intelligence, is one of the most striking developments of Ukraine crisis, close quote. Between the intel failure, if you're going to grant that it was a failure, in Afghanistan last summer, and the intel success in Ukraine this spring, what changed?

Amy Zegart: Oh, such a great question. So I'd take issue with the premise, Peter

Peter Robinson: Would you?

Amy Zegart: that Afghanistan was an intelligence failure. So as you know, there's an old saying in intelligence business that there are never policy failures, there are only intelligence failures. And so my read of Afghanistan was it really was a colossal policy failure. And in fact, if you take a closer look at different intelligence agencies and what they said, the CIA was always more pessimistic about the ability of the government in Afghanistan to sustain itself than the military was. I think part of the challenge there was defense intelligence agencies were grading their own homework. So it's certainly true that they were more optimistic than they should have been. But the reason that withdrawal was so devastatingly unsuccessful had less to do with our intelligence agencies, and much more to do with the Biden Administration's policies. You know, intelligence is a moving target. We can't say the government in Kabul is going to sustain itself for X number of days, when what we are doing on the ground is affecting how the government of Afghanistan is going to sustain itself. So it's an interaction between policy and intelligence. And so I think it's really a policy failure in Afghanistan.

Peter Robinson: Okay, so if you don't, I grant that you're rejecting the premise, do you still see, I guess the question is, did the intel community, the agencies, this vast apparatus that you're going to explain to us, was there some real time learning in these last six or eight months? Or did they get it right the first time, and they got it right this time, and that's just being professional, consistently professional throughout the period?

Amy Zegart: I think we don't know. I think it's gonna take a long time before we're going to know why we did so well with Russia's invasion of Ukraine. You know it depends on the time. It depends on the topic. Some topics have better talent, some topics have better access. What is clear about Ukraine is we had really good intelligence. We had really good collection, it looks like from multiple types of intelligence, human intelligence, technical intelligence. We had great analysis of that intelligence and what it could suggest. And perhaps most importantly, the administration used the intelligence, it declassified the intelligence in real-time, and that's new.

Peter Robinson: All right. Okay. Oh, I see. So it's the policy people. If there was learning that took place, it's the policy people who did a large portion of that learning.

Amy Zegart: I think that's one big difference between Afghanistan and Russia, which is the policy people actually telegraphed, here's what we're learning in real-time about troop movements, and false flag operations, and letting the world know Putin's trying to con the world. Don't believe the lie. We're getting the truth out first.

Peter Robinson: Okay, so here's the underlying theme, I hope, that emerges in our talk, because it's the theme that popped up in my head again and again as I read your book, and that is the difficulty of operating an intelligence community in a democracy. We pay their budgets. Their budgets are dark. We don't really know much about the budgeting. They get to operate in secret. And then you have this persistent problem, this leads to the next two or three questions, that we don't know what the baseline is. How do you know when they got something right? And when they got something wrong? And if you don't know that, how can you hold them accountable? Okay, so that's sort of the underlying, or the background question. And if you don't mind, I'd like to take you through a couple of case studies, and just ask how a professional, and a scholar, such as yourself, how you think about these questions, all right?

Amy Zegart: Okay.

Peter Robinson: So we started with Ukraine, which is of the moment, but let's back up. Iraq and the weapons of mass destruction. Meeting in the White House in 2002, CIA director George Tenet tells President Bush, the evidence that Iraq possesses weapons of mass destruction amounts to a quote, a slam dunk case, close quote. And of course, we invade the next year, and WMDs aren't there. All right. After the war, the Robb-Silberman Commission investigates what happened. They investigate the intelligence, particularly. And the central finding, which judge Larry Silberman discussed at this very table a few years ago is this, Saddam Hussein had persuaded everyone in Iraq, his generals, his family, that he did have weapons of mass destruction. And if everybody believed something that's untrue, whether it's untrue or not, it's very, very hard for intelligence to pick that up. Do you buy all of that?

Amy Zegart: I buy some of it. I think, you know, we have access now to the Ba'ath Party archives. We have access to a lot more intelligence about how Saddam dealt with his inner circle. And it's clear that he was duplicitous even with his own people.

Peter Robinson: Right.

Amy Zegart: And he had reasons to do that. He wanted the world to think he had these capabilities at the same time he wanted the Americans to believe that he didn't. And so that part is certainly true. But it's also true that we had colossal collection failures of intelligence.

Peter Robinson: Okay, explain that. What is a collection? You use that moment ago when you were saying we had good collection

Amy Zegart: Right.

Peter Robinson: for Ukraine. What does that mean?

Amy Zegart: So good intelligence really requires two things. You have to collect the right information and then you have to assess what it means. And information is often ambiguous, so you have to have both things. And in the case of Iraq and WMD, we really didn't have people on the ground after weapons inspectors were kicked out of the country. And so our information coming out of Iraq was really frozen in time. And so what that then led to was a series of analytic errors. In the absence of new intelligence, our intelligence agencies assumed Saddam was doing what he had done before. When in fact he wasn't. And so that commission and other commissions that investigated the WMD failure really found a number of different analytic problems. Group think being one of them. Nobody really took dissenting views seriously, for example. So we know that we're dissenting views in the State Department, and in the Department of Energy, and their intelligence unit about whether Saddam really had the capabilities that we thought he did. There were dissenting views, but they were buried in that national intelligence estimate. Well, we also know that there was sort of mirror imaging. We imagine that Saddam would behave, as we would behave, in that situation. And that turned out not to be true, too. So as political scientists like to say, the failure in Iraq was overdetermined. There were so many variables that went wrong. Lots of factors that went into this mistake. But it was clearly an intelligence failure.

Peter Robinson: I see. All right. Here's another case study. And you go into that one in your book, of course, and you go into this one in your book as well, and that's the search for Osama bin Laden. Very briefly, and as you explain in the book, the hunt for bin Laden took a decade. And for most of those years intelligence officers told us, at least this is me speaking as a layman, having read the newspapers, and paid some attention to it over of years, that bin Laden was most likely in this very difficult, mountainous terrain between Afghanistan. You're smiling because you heard this a thousand times yourself, and

Amy Zegart: Everybody was talking about this for years, right.

Peter Robinson: And when we found him, nothing of the sort. He was living in a comfortable compound, less than a mile from the Pakistani military academy. He wasn't in the mountains. He was in a perfectly inhabited part of Pakistan. So assessing, how do we assess this? On the one hand, we found him, and we got him. On the other hand, it took a decade. And who knows how many tens of millions of dollars to find a man who was, well, not exactly hiding in plain sight, but he certainly wasn't up in the inaccessible mountains. So how do you, I guess what I'm asking is, how does an ordinary American, how does a member of Congress who sits on an oversight committee, but always has to go back to his office and make fundraising phone calls, or go home to a campaign in these people are busy people. They're seldom experts. Members of Congress are the people we get. How do you advise them on how to weigh, was the search for Osama bin Laden a success or a failure? Should it have been quicker? How do we evaluate it?

Amy Zegart: Such a hard question.

Peter Robinson: Oh, it is a hard, okay, well, good. I'm happy. Next question. Seriously, that's one thing I wanna hear is, is that harder, or to a professional, is it easy?

Amy Zegart: No, it's really hard. So I remember years ago, I was in Los Angeles when then CIA Director Michael Hayden came to speak.

Peter Robinson: Right.

Amy Zegart: And we had not found Osama bin Laden. And he gave this public speech. And the first question from the audience, gentleman raised his hand and he said, I wanna know why after all the billions of dollars we spend on these 18 different agencies, however many there were, all these different intelligence agencies, why can't we find Osama bin Laden? And Hayden shot back without missing a beat, he said, I'll tell you why, because he's hiding. And everyone started to laugh. But Hayden's point was really serious, which is that this is hard stuff. It's a lot harder than we think it is when we watch television and the movies. So I look at the bin Laden operation as an intelligent success. And one of the reasons I look at it as an intelligent success, is if you look at, you noted before, Peter, that we all thought bin Laden would be hiding in the mountainous regions between Afghanistan and Pakistan. Why? that's what he had always done before. His trade craft, if you will, how he dealt with his own security to keep from getting discovered, was to hide out in mountainous regions, to surround himself with security forces, to not communicate via satellite phone and other things, and to stay separate from his family. That's how he had always done it for his entire life. And guess what? Osama bin Laden changed it all. He zigged instead of zagged. And so he ended up hiding in an urban area. He ended up hiding with very little security. He ended up hiding surrounded by family members. So for the CIA to find him, they had to discard all sorts of assumptions that were very reasonable assumptions that very smart people would make about how bin Laden might be operating. And not only did they have to do that, they had to pick up the clues that were dispersed across time and place. And so the fact that we ultimately did find bin Laden is actually an incredible success story. And I think the intelligence community, we often talk about its failures. We rarely talk about it successes, in part because they remain classified.

Peter Robinson: Yes.

Amy Zegart: I think this is an example of an intelligent success.

Peter Robinson: All right. An entirely new kind of threat. I'm talking about Osama bin Laden. People can picture him. He's a human being. We get him by sending guys in on helicopter, okay. Now here's something that's intangible and new. And that's the cyber threat, to which you devote a couple of chapters in Spies, Lies, and Algorithms. Let me quote the book, quoting you, put simply, cyber attacks involve any activity, any activity, that alters the confidentiality, integrity, or availability of information on digital systems. Okay. So the bad guys can and use cyber attacks to steal information from corporate, ranging from corporate plans to nuclear secrets, to blackmail high officials, snooping around in their online garbage bin, so to speak, to threaten to shut down portions of the grid. In an extreme case, a cyber attacks could make it possible for our planes or ships to navigate. Could really shut us down, right?

Amy Zegart: Yes.

Peter Robinson: I'm not overstating the case.

Amy Zegart: No.

Peter Robinson: I'm reading your book correctly on all this. Okay. So this is bloodless by comparison with a nuclear attack. I'm just trying to get the magnitude of the threat here, but cyber attacks could represent as grave a threat to our security as nuclear weapons. Am I being melodramatic or not?

Amy Zegart: No, I don't think you're being melodramatic at all.

Peter Robinson: Okay, so question number one. How good are the bad guys at using cyber attacks already? So what I'm asking for here is some kind of assessment, some way to convey how much damage we've already suffered at the hands of, say, the Chinese and the Russians, because of cyber attacks.

Amy Zegart: Well, I think, Peter, you know, we off often talk about war, and how do we know when we're in cyber war? And I think we already are, most Americans just don't know it. So we're being attacked millions of times a day. The Chinese have been still

Peter Robinson: You say this is though it's blah, but

Amy Zegart: Well, in the national security world,

Peter Robinson: To you, it is.

Amy Zegart: it is an everyday thing. And I think one of the challenges that you raise with cyber versus nuclear is most Americans think of cyber attacks as my credit card got stolen, it's kind of an annoyance.

Peter Robinson: Right.

Amy Zegart: And it's something that's intangible. It's not that serious. Maybe it could result in a shutdown of a company for a day or two, but it's hard to fathom how cyber attacks could operate, because we haven't seen the results of a massive cyber attack that shut down our economy for example before. It's not to say it can't happen, but we haven't seen it. So it's hard to imagine what it could be. But the reality is that there are major cyber actors that have been waging cyber warfare against this country for years. I put China at the top of that list. China has stolen us blind. China has stolen billions and billions of dollars of trade secrets and intellectual property.

Peter Robinson: As a matter of state policy, as a matter of policy, Chinese Communist Party, not this or that individual corporation, it's not Ali Baba

Amy Zegart: No.

Peter Robinson: that's stealing secrets from Google.

Amy Zegart: No, this is the Chinese Communist Party, either tacitly supporting or actively directing

Peter Robinson: Right.

Amy Zegart: cyber espionage and theft of American trade secrets for technological advantage.

Peter Robinson: Right.

Amy Zegart: And they've been doing it for years.

Peter Robinson: And that is the action of an enemy is the way we ordinarily, I mean, there's all, when we think of China, are they a competitor, an adversary, or an enemy? You've just described the activities of an enemy, haven't you?

Amy Zegart: Yes.

Peter Robinson: All right.

Amy Zegart: And I think the thinking in this country has changed. I think the Trump Administration to its credit really highlighted the fact that China is not a responsible stakeholder, and that China really is pursuing interests that are against freedom, they are against the United States, and they're against the world order. But, you know, there are senior military officials that have said, every major Chinese weapon system today is based on technology stolen from the United States. Every major one. So we think about, not just on the commercial side, but in the military side, why is China's military getting so good, so fast? Stolen intellectual properties Stolen technology from the United States. So they're spying and they're stealing. And then they're the Russians, which are, you know, the Russians are very good at deception operations.

Peter Robinson: Right.

Amy Zegart: And so they've been very active in information warfare in this space. And of course, in ransomware and these attacks that hold companies and individuals hostage. And so then there are two other major cyber actors that I put in the big four. The other two are North Korea, surprisingly adept at cyber capabilities, and Iran, which has waged destructive attacks against casinos, against American banks, against the Saudi oil company. And so when we think about the cyber threat landscape, these big four, China, Russia, Iran and North Korea have a lot of things in common. Number one, they're sophisticated cyber adversaries. Number two, they all either have, or want to have nuclear weapons. Number three, they all are seeking or are in the midst of pursuing territorial aggression in their neighborhoods. Russia's invaded Ukraine. China has its sights set on Taiwan and the South and East China Sea in contested territory there. North Korea, South Korea, and Iran in its neighborhood. And then all four of them seek to remake the international order in a way that is not good for democracies and freedom loving people. And it's that combination of factors, the cyber with the nuclear, with the territorial aggression, and the remaking of the international order, it's that combination that makes those four countries in particular so dangerous for the United States.

Peter Robinson: Okay, do you know how depressing all that sounds? You're speaking in a sophisticated, nuanced way, the professional academics stating threats that you now are so, but I read your book, and then I hear you say this for, you might as well be Darth Vader breathing heavily.

Amy Zegart: So this is why my husband says I can't talk about what I do at cocktail parties.

Peter Robinson: Right.

Amy Zegart: Because it depresses people.

Peter Robinson: People start drinking too many, too quickly.

Amy Zegart: Yes.

Peter Robinson: All right, so how good are we at facing up to these threats? You write about the distinction within the intelligence community is people who pride themselves on being pure intelligence officers or officials, and those who view themselves as war fighters instead. And these are two different cultures within intelligence. And you write about this as a problem for standing up, in particular to the cyber threat. Could you explain that? And also can, I'm sorry, I'm just gonna give you a sort of big question then just drop it in your lap. The other piece of this is, where do we stand in regard to these cyber threats, which the Chinese have already stolen us blind? That's what you said. They've stolen us blind. Is this Pearl Harbor, where we're just finally waking up to threat, or is this Guadalcanal? Have we won some major encounters? Do we know what's going on and we have an apparatus at work? All right, all of that, Amy.

Amy Zegart: In 30 seconds or less.

Peter Robinson: With the further assignment that the end has to be cheerful.

Amy Zegart: Okay. I actually think the end is cheerful.

Peter Robinson: All right.

Amy Zegart: Let me start with the good news. How about that?

Peter Robinson: Thank you. Thank you.

Amy Zegart: We are getting much better in cyberspace. So we now have Cyber Command. That's about a decade old. So it's important

Peter Robinson: And it's located where, in the Pentagon.

Amy Zegart: It's shares space in Fort Meade, so this is important, so

Peter Robinson: Right.

Amy Zegart: The National Security Agency and U.S. Cyber Command are very tightly intertwined. There's a big debate about when they should go there separate ways. And the reason this is important is that cyber capabilities descend from an intelligence bloodline. So the National Security.

Peter Robinson: That's the pure intelligence side.

Amy Zegart: Yes. So the National Security Agency collects foreign signals intelligence. So think intercepting phone calls, emails, encrypted communications. I emphasize the word foreign, right? They're not spying on your phone calls with your grandma. So that's their job. And so cyber capabilities descended from that enterprise. And so the head of the National Security Agency has, until now, always been dual hatted as also the head of United States Cyber Command. So this question of intelligence versus war fighting is very much at play. Intelligence officers always want to keep collecting. They work hard to penetrate adversary computer systems, and they don't wanna give that up by revealing what we know and using it, right? Either to defend or to attack. So intelligence officers always want to err on the side of keeping the listening going, keeping the intelligence collection streams going. War fighters have a bias for action. They wanna use that intelligence to do something. And so you have these two tendencies co-located in the same, well, not the same organization, but a highly overlapping organization. And so what we've seen over time is a move from, we're gonna really have an err on the side of understanding, and listening and defending, right? This perimeter defense sort of perspective of cybersecurity, which didn't work out so well. To, we need to be more aggressive. We need to be more forward leaning. We need to go on cyber offense. And so what we've seen in the past few years is a really new strategy, which Cyber Command calls Defend Forward. And as its name suggests, it means we can't just sit here and defend against attacks on our networks. We have to take the fight to adversary networks. We have to impose friction on them. We have to make it harder for the bad guys to operate in their own computer systems. And that's a better way to defend. And so we've gotten much better at that in cyberspace. And I think you can see sort of tell-tell indicators of that in the comments that we're hearing from the Commander of Cyber Command, General Nakasone, in Congress last week alluded to the fact that we've been helping the Ukrainians defend their cyber capabilities better. I was listening to a wonderful event on Zoom last night, with five former directors of intelligence agencies and General Hayden, who used to run the National Security Agency, said something to the effect of, if we look back 10 years from now, we're gonna be pretty happy with how our cyber capabilities were used. Suggesting that in fact, we've been more active than maybe the government would say publicly. So that's a good news story. We're getting better. And we're getting more assertive at using our tremendous cyber capabilities to Defend Forward. But it's a changing battle space, and it's a really complicated and challenging battle space.

Peter Robinson: Okay, you said you were gonna start with the good news. The bad news is it's complicated. Oh, there's more bad news. Well, we'll just wait for that. All right, listen yeah, go ahead.

Amy Zegart: Let me just put a sort of a framing around the bad news. Why is cyber so hard and why should we be so afraid?

Peter Robinson: Yes. Yes.

Amy Zegart: I think there are two reasons. If we think about what historically has protected the United States from all sorts of bad actors, what makes us secure as a country?

Peter Robinson: Water

Amy Zegart: Water. So geography is one important one. So we're physically separated from bad neighborhoods in the world. And the other is power, right? Our military power in particular. The most powerful militaries can better protect their societies. Those two things do not protect us in cyberspace. So there are no good or bad neighborhoods in cyberspace. There all bad neighborhoods in cyberspace. That's how the internet was created to be open. So you can't protect yourself no matter how much you try. You can, oh, I should take a step back. You can't fully protect yourself, right? It's not like the oceans. We can't create cyber oceans. And then

Peter Robinson: There's no coherent isolationist impulses in cyberspace. You just can't do it.

Amy Zegart: And there's a trade off. The more you isolate from cyberspace, the more you hurt global trade, the global exchange of ideas. So there's an inherent trade off there. So there's the Great Firewall in China, right? But it's not as much of a firewall as the Chinese Communist Party would like it to be, number one. And it has a real price to pay in terms of the ability of people to communicate with each other, number two. So that's the first thing that historically has protected us, is water. But the second, is power. And the United States is one of the most powerful countries in cyberspace and simultaneously one of the most vulnerable countries in cyberspace. And that's new. That's unique to cyber. Because we're asymmetrically dependent on all this connectivity for our economy, for our society, for our government. And for our freedom of discourse. So because we're a democracy, because we value free speech, it means that bad guys can deceive at scale. So we're asymmetrically vulnerable to a host of cyber bad actors that we aren't in physical space.

Peter Robinson: All right. We'll come back to that.

Amy Zegart: See, I'm trying to depress you a little bit.

Peter Robinson: Yeah, no, no, you're good at that. Intel in Silicon Valley, Spies, Lies, and Algorithms, I'm quoting you, at the exact moment when harnessing technology is the key to success, it's hard to overstate just how foreign the worlds of Washington and Silicon Valley are to each other. You inhabit both worlds. Or you've served in high positions in Washington. You're in touch with intelligence in Washington. And yet you live right here at the epicenter of Silicon Valley. What do you mean?

Amy Zegart: So I call it the suit-hoodie divide. So, you know, here in the Valley, people even dress differently than they dress in Washington. They speak a different language. They have a different culture. There's a different orientation. That's not to say it's unbridgeable, but it is in many ways bringing together two foreign cultures. So I have said this to my friends in the Pentagon, stop coming to Silicon Valley and using all of your D words. Military officials love to come here and they use words like, destroy, dominate, degrade, defeat. These are very popular words in the military. They are terrifying words in the Valley, In the Valley, they like to use C words, create, collaborate, change, culture. And so even if they're aligned interests to protect the nation or advance the national interest, it's often hard for these two cultures even to understand each other to find common ground.

Peter Robinson: Once again, from Spies, Lies, and Algorithms, quote, when Facebook's CEO Mark Zuckerberg testified before Congress in April, 2018, it was a jaw-dropping moment showing just how much profits were driving Facebook's decisions, and just how little Zuckerberg and his team had considered the possibility that anyone could use their platform to undermine democracy. Google executives canceled an artificial intelligence project with the Pentagon, and refused to bid on a desperately needed $10 billion program to modernize the military's cloud computing program, close quote. Okay, I read that, and my first instinctive response is, Facebook and Google should be ashamed of themselves. And then my second response kicks in, wait, just a moment, their employees and their revenues are completely international. They both employ lots of Chinese, Chinese nationals. They have offices all over the world. Why should they feel any particular allegiance, let alone duty, to the United States?

Amy Zegart: So I completely agree with you. You can see why this is such a fraught business. Because these companies have global shareholders, not just global employees and global markets, and they have global interests. But we also have national security interests, and they're American companies, and they need to have American responsibilities. So how do you find the right balance between global corporations and national security? And that's what we're seeing, sort of both sides trying to maneuver to find more common ground. It's a very challenging set of relationships. You know, I took a look at Google's AI ethics, for example, and they say we never wanna be involved in anything that could be a weapon of war. And my response is, you are a weapon of war.

Peter Robinson: Yes.

Amy Zegart: Google is a platform by which adversaries do all sorts of bad things. So you can't opt out of geopolitics if you're a company and that's especially true for tech platforms. They don't, and I often joke that government agencies want technological capabilities from the Valley that they don't have. And Silicon Valley companies have responsibilities that they don't want. And so each side is grappling with these really difficult challenges about how to be responsible for national security. I don't think people in Google are bad people with evil ideas, but I think there's a naivete among these tech companies that they don't need to think about negative use cases of their technologies, and that's not their problem. It is their problem.

Peter Robinson: Could I, again, this is wonderful because I read the book and now I get the chance to tell you some of the things that came to mind as I was reading. And I read this passage about Zuckerberg and the current people at Google and so forth. And I think to myself, I think to myself of the old Silicon Valley, which was still around when I first came out here. General William Draper founded, I think, the first VC firm is generally considered to have been the first venture capital firm. His son, Bill, and now his grandson also venture capitalists. But listen to a little bit about the original William Draper. He served in the Army. He served as Undersecretary of the Army. He served as Ambassador to NATO. David Packard. Mr. Packard was still on the board of the Hoover Institution when I first came here. He founded Hewlett-Packard. And what else did he do? He served as Deputy Secretary of Defense. So my point here is, that the generation that founded Silicon Valley were figures of undoubted patriotism who understood the need to, I don't wanna say cooperate, but they somehow or other wouldn't have occurred to David Packard that he was in any business other than an American business, that he owed something to this country. Okay, so I think why should that have been the case? And what I can come up with, this is just what occurs to me, I wanna see what you make of it. General William Draper is called General for a reason. The founders of Silicon Valley had experience of war. They understood the stakes. They knew what could happen if things went wrong. If our position eroded relative to that of our enemies, and the current generation, I'm a little suspicious of this thought, because it's so facile, which is why I'm putting it to you, but the current generation, these people all got rich remarkably quickly, and without paying any price, without seeing the stakes, without understanding what could go wrong, if things do go wrong, does that make any sense to you? And if that's the case, what do we do except wait for another massive global confrontation to teach them their rotten lesson?

Amy Zegart: So I think that's a big part of what we're seeing. So if you had asked the question a few years ago, fill in the blank, China is a blank of the United States, and you had asked that to people in Washington, you would've gotten competitor, perhaps adversary, enemy. If you had asked that question here in the Valley, you would've gotten market, opportunity.

Peter Robinson: Right.

Amy Zegart: Investor.

Peter Robinson: Right.

Amy Zegart: So it's just a different perspective. I think that's different now. I think the China threat's much more obvious. I think the gap, the divide between Silicon Valley and Washington is narrowing. I think it's getting

Peter Robinson: You do?

Amy Zegart: I do. But I think there's something else in what you said, Peter, which is that, you know, this generation, not only did they not serve these founders of these companies, they don't know anyone who did. And so they're living in different universes. How many people in these tech companies know anyone who's ever been in the military before? And conversely, how many people working in the Pentagon know people who grew up in the Valley, and what their ethos is, and what they value? And I'd layer one other thing into this conversation, which is engineers. Engineers like to solve technical problems, and they don't think they're in the policy business, right? So one of the great things about being at Stanford is I get to teach a lot of engineering students.

Peter Robinson: Right.

Amy Zegart: And I wanna encourage them to about international security as part of what they do. 'Cause I want the Mark Zuckerbergs of tomorrow to walk in the door the first day, better understanding the potential negative uses of their technology. And so we're seeing a slow movement to understand that engineers are not just engineers, they are policy makers, and the design choices they make, make it easier or harder for bad people to do bad things with their inventions. And so you can't outsource policy to the Hoover Institution and people like me, engineers have to think that policy is their job too. And they have to bake it in to their engineering decisions. We're getting there. But I think that's a big sea change that needs to happen, not just in the private sector environment, but in universities as well.

Peter Robinson: I see. Okay. All this is fascinating. Back to this question, this sort of special problem of intel in a democracy. Let me tell you a horror story. It's a true horror story. I mentioned Judge Silberman a moment to go, who was with former Senator Robb chaired the Robb-Silberman Commission. Early in his career, he was at the Department of Justice at that point. And this was in the early seventies when there was the Church Commission, which you write about in Spies, Lies and Algorithms, was discovering bad behavior by the CIA, and J. Edgar Hoover had left the FBI, and it fell to Judge Silberman to read through J. Edgar Hoover's private files. And he wrote about this experience, he published a piece in the Wall Street Journal in 2005. This is years after he read through the files. It was the single worst experience of my long governmental service. Hoover, J. Edgar Hoover, had indeed tasked his agents with reporting privately to him any dirt on figures, such a as Martin Luther King or their families. Hoover used that information for subtle blackmail to ensure his and the bureau's power. Perhaps even worse was the evidence that Hoover had allowed the bureau to be used by presidents for nakedly political purposes. And Silberman notes, that LBJ in particular was a bad actor. He and J. Edgar Hoover were using information for nakedly political reasons. Judge Silberman concludes, this country would be well served if the name of J. Edgar Hoover were removed from the FBI building, close quote. So Larry Silberman, not a liberal, tough guy, brilliant judge, broadly speaking a conservative, and he was disgusted. Now, I'm putting it back with J. Edgar Hoover, and I'm avoiding the whole set of controversies about Donald Trump, and Russian disinformation, and Hunter Biden's, let's just put that away, because it's almost impossible to talk about that without getting drawn into other, the question here is, and I think Hoover, Hoover may be an extreme case, but he's a case, he ran the FBI. They get to operate in secret. We don't even know much about their budgets except round numbers, and you yourself write that congressional oversight is terrible, just terrible. So just this kind of threshold question I've already asked, how do we assess whether they're doing their work as they should? How can we trust these people? I'm putting that too broadly. But how can you trust any human being when you say you get to operate in secret? We're going to give you the immense power that comes from secret knowledge. We're going to give you budgets about which very few details will ever be requested. Now go behave in a noble, idealistic and completely selfless way. That's just not the way human beings operate. Not all of them. So how can we trust these people?

Amy Zegart: Well, you've put your finger on a conundrum of secret agencies operating effectively in democratic societies. We have always grappled with this. And the J. Edgar Hoover years were many. He ran the FBI for decades.

Peter Robinson: Yes.

Amy Zegart: And they were a dark chapter in American intelligence history. And the FBI wasn't the only agency that violated civil liberties that was used in political ways, that read Americans' mail, that did dirty deeds in the case of the CIA abroad in terms of assassination plots. And so the question is, how do we know? How can we make sure these agencies are accountable, that they're working according to the law and our values? Congress has to play a crucial role. So before the seventies, as I write in the book, there there were no permanent oversight committees. Oversight was sort of, we don't really wanna know in Congress, and we're not gonna ask. And maybe we'll just spend a few minutes a year. And so that's better than it used to be. Now there actually are oversight committees, and they do hold hearings, but it's nowhere near what it should be. These oversight committees play a pivotal role because they do get the classified briefings. The intelligence agencies are required by law to keep them fully and currently informed. And Congress has tools at its disposal to punish them if, in fact, they don't keep the committees fully and currently informed. They do have the power of the purse after all, among many other things. And so the question is, how can they do that job better? And the problem is, you know, I'm a political scientist, so I look at incentives and institutions.

Peter Robinson: Of course, right, right, right.

Amy Zegart: And so I think what we often hear, Peter, is the narrative is well, oversight is better or worse depending on who's president, or who's in charge of the committees. And what I find is actually oversight is almost always worse than it should be, because of incentives and institutions. Voters don't care how their member of Congress works for intelligence oversight. Oversight service and intelligence is kind of an electoral loser, right? They can't even talk about what they do to the folks back home. And so unless you wanna of burnish your national security credentials, 'cause you're thinking of running for president, most members of Congress are not drawn to serve on these committees to begin with. And they're not experts. So one of the things I found in doing research for this book, is there are more dairy experts in Congress, right, or powdered milk experts in Congress than intelligence experts. Why is that? Because there's so many dairy districts where members of Congress have to learn the business if they wanna get reelected. There's no dairy district for intelligence in the same way. And so that means members of Congress have to learn on the job. And learning about intelligence takes a lot of time. And you have to do your homework in a secured facility away from your office without your staff. And if it's hard to do your homework, the homework doesn't get done. So the key here to get better oversight is voters have to care. Voters have to tell their members that they care. Voters have to incentivize members of Congress to care more and we'll get better oversight from them. What we've seen in the past few years though, Peter, is some outsourcing that can be useful of intelligence oversight to a couple of different types of institutions. So one is inspector generals. So these are people who are answerable to Congress and they can actually examine wrongdoing inside the agencies that they're overseeing and we

Peter Robinson: Better than nothing, but they only get called in when something's already gone wrong.

Amy Zegart: Correct.

Peter Robinson: Isn't that right?

Amy Zegart: Right. Yes. And so when we think about, what's the proactive effort to make sure these agencies are effective, not just accountable? We need both. Congress has to do that and they don't do a very good job. Congress is very good at jumping into scandals after they've happened to complain about how awful things are, when in fact they need to stop the fire from breaking out in the first place. They should be working in partnership with our intelligence agencies.

Peter Robinson: I would seem to me, I'm not a political scientist, and I'm acutely conscious that I'm talking to a very good one. I'm not even sure I'll frame this question well, but it seemed to me that presidents have a big incentive to make sure the information that they're being given is pertinent and correct. And, of course, with the case of J. Edgar Hoover and LBJ, J. Edgar Hoover and LBJ are, so to speak, in bed together. Set that aside. Presidents don't want intelligence agencies snooping around, George W. Bush, you'd have to say, you yourself said there was a colossal or you used some word, I don't know, there was a big intelligence failure in Iraq. Who more than anyone else, excuse me, the people who died as a result, of course, we're the ones who suffered the most, but in political terms, surely it's George W. Bush and Dick Cheney, whose reputations will take decades to recover, if indeed they ever do, right?

Amy Zegart: Yeah.

Peter Robinson: And now, so don't don't administrations, aren't they in the position, I'm conscious that George W. Bush kept a professional in the job. George Tenet had served under Clinton and George W. Bush kept him on. Whereas, Ronald Reagan, my old boss, no, thank you. I want guys I know running CIA, Bill Casey and at FBI Judge Webster. I don't know. Isn't that the way the system ought to work? Isn't it presidents who ought to keep tabs on these guys

Amy Zegart: Well, we

Peter Robinson: who have the incentive to do so?

Amy Zegart: Presidents do, but we had real problems in the Reagan Administration, too, with Bill Casey running CIA, with Iran-Contra, for example. So the CIA got over at skis and some covert action that proved problematic for the Reagan Administration politically. So Reagan suffered politically too because what the CIA was doing. So, but you're right, presidents.

Peter Robinson: I'll stick up for Bill Casey, but that's a separate show. But the general point that things got became a mess. Yes, I grant that. Go ahead.

Amy Zegart: So presidents have an incentive to make the intelligence community work well for them. The problem is they don't have the time. Presidents have, as you know, very short time horizons, and they have long agendas. And those agendas are about action. They're about policy. They're not about delving into the bureaucratic details of whether the agencies are working as they should. And so they're focused on the near term, which means that these long term weaknesses, are we adapting to the end of the Cold War and the rise of terrorism? Are we understanding what's going in cyberspace fast enough, well enough? Those longer term challenges almost always get put lower down the priority list for presidents. And by the way, when presidents say they do wanna actually have the intelligence agencies work better together, they're resisted, right? The most powerful interest group in Washington. is the status quo.

Peter Robinson: Correct. Correct. So I don't wanna stick with this forever, but if polls are to be believed, it's very likely, a lot can happen between now and November, but it's very likely that the next Speaker of the House will be Congressman Kevin McCarthy from the Central Valley here in California. So what do you advise him? You're very likely to become Speaker, here's a really serious problem. And if you do become Speaker, it'll be under circumstances in which the whole world is expecting new departures. Here's what you should do.

Amy Zegart: I think with respect to intelligence, I'd say, choose the members of that committee carefully so that they have a bipartisan ethos, that they think about the nation. I think we've seen a lot of this on the Senate side, frankly, they've done a really good job

Peter Robinson: Of being bipartisan

Amy Zegart: of remaining bipartisan in a very polarized time when they have to, to support the intelligence community to do a better job. That's number one. And number two, and I've mentioned this to some members of the committee already, if you had to do one thing immediately that would improve oversight in the House, get rid of the term limits on the house intelligence committee. So if you think about

Peter Robinson: It takes time to get to know the community,

Amy Zegart: Yes.

Peter Robinson:  the procedures, the protocols.

Amy Zegart: and just when you learn what all those acronyms mean, you have to get off the committee.

Peter Robinson: I see. Okay. Last questions. I began with a couple of case studies set in the fairly deep past, at least by the standards of students, let's say, 10, 12, 20 years ago, was a long time ago.

Amy Zegart: That's like the Peloponnesian War for undergraduates.

Peter Robinson: Yes, exactly. That's exactly right. That's exactly right. So let's start with one set in the present, or the near future. China has stepped up its overflight of Taiwan. I've even read commentary, this is going on all the time. Fighters are violating Taiwan airspace all the time, Chinese fighters. And the speculation that I've read is that they're trying to numb the Taiwanese, that they're trying to make this, this is a first step to broadening the possibilities for what they might do. And step number one is, you make violations of their airspace feel routine even to them. So now that Russia has invaded Ukraine, we need to know from the intelligence communities, what are the Chinese thinking about Taiwan? How do you frame the question? What expectations do you set? How long should it take to get useful information? So you're advising the president. Let's say you're advising Joe Biden on that very set of, how do you task the intelligence community and what are the reasonable expectations you can have for them?

Amy Zegart: So I think, Peter, there's the baseline question of capabilities and intentions.

Peter Robinson: Right

Amy Zegart: So let's look at that before the invasion of Ukraine, what does Xi Jinping want? Well, we know what he wants.

Peter Robinson: He tells us.

Amy Zegart: Hell tells us what he wants.

Amy Zegart: Yes. Yes. Right, he wants reunification with Taiwan, ideally through coercion, not through military action. But as you pointed out, you know, the Chinese are getting much more assertive in using their newfound military power to desensitize the Taiwanese and us, by the way, about what they're doing. You see this build up in the South China Sea. So they have dramatically increasing capabilities and we know what their intentions are. So that's the baseline. How does the invasion of Ukraine change either of those things, how the government views its own capabilities? because it's clear the Russians, and Putin in particular, overestimated his own military and what it could do. They hadn't really fought in a contested environment in years. And the Chinese Army hasn't fought in a contested environment in longer. So if I'm Xi Jinping

Peter Robinson: Since 1949.

Amy Zegart: Since, or '79 with Vietnam.

Peter Robinson: All right, fine.

Amy Zegart: So if I'm Xi Jinping, I'm a little worried about is my military as good as I think it is? And are the capabilities that I've been developing really as battle tested as I want them to be? I'd be a little more concerned about that. So I would wanna know from the intelligence community, what are we picking up about what the government is reassessing in its own military? Then there's the question of what are his intentions in, what are the lessons that the leadership in Beijing is learning from this invasion? And it's too soon to tell, but it's clear over the longer term that Vladimir Putin has United the west in a way no one thought would be possible. That the U.S. led international order's gotten a shot in the arm. That the allies have come together. And that we're using economic tools in ways we have not used before. And if you're Xi Jinping and you've bet the farm on your regime being tied to economic growth in China, between COVID, and what you're seeing with economic sanctions, you gotta be a lot more worried about a Taiwan scenario. Now, that said, I think it's fair to surmise that the world would not respond to China with sanctions in the way it's responded to Russia, because we're so much more tightly intertwined economically with China. So we would feel much more pain, not just the Chinese. But I think from an intelligence perspective it's, what's the baseline? How, if at all, do we think this might be changing in which direction? And how would we know it when we saw it? How do we know what the thinking and what the capabilities are in China as these events play out? What lessons is Xi Jinping drawing from what he's seeing on the ground? And, by the way, if I were queen for a day, how can the United States use this horrific invasion as a geopolitical opportunity to drive a wedge between the Russians and the Chinese?

Peter Robinson: Ah, no, that's a policy question.

Amy Zegart: It is. But there's an intelligence piece to it. Which is, what are the conversations happening between Moscow and Beijing?

Peter Robinson: I see.

Amy Zegart: What are the points of friction between the two? What are they negotiating over? How can we publicize what we know? You'll notice there's been declassified intelligence about Russian requests for Chinese assistance.

Peter Robinson: Right.

Amy Zegart: I think that's pretty smart. How can we publicize what we know to back the Chinese into a corner? They have to take a side, because they're in a very uncomfortable position, and Russia is not the partner you want right now in the world. So I think there's an opportunity to use intelligence and disclose intelligence to support an exploited divide between this growing alliance, this allegiance between the Russians and the Chinese.

Peter Robinson: Okay. Another couple of questions, if I may, these last questions. China's huge. Outnumbers us by five or six to one. Its economy is still growing. It's capable of tremendous growth. And it's ruthless, centrally controlled. Okay. Russia is vast, geographically, of course, that represents an advantage in some ways. What you have in Russia is a willingness to use whatever, it has a weak hand, declining population, but Putin is willing to pull triggers. All right. What advantage do we have, long term advantage with regards to intel? We're only 330 million people. We're politically very bitterly divided, at least at the moment. Hard to imagine that turning around quickly, maybe single digit years, but maybe not. And you've already said that our capacity for technical innovation as an advantage that's at least muted by the Chinese ability to steal innovations two seconds after they occurred to some bright engineer at Google or Facebook. So why aren't we doomed? What are the advantages we have with regard to intelligence that are distinctively American and that give us some hope? What can we do that they can't counter? What can we do that they can't do? Or that they can't do nearly as well?

Amy Zegart: I think, Peter, we have some real enduring advantages in intelligence. The first, is we have allies. You know, China has customers, it doesn't have allies. And what we're seeing this in Ukraine, too. So we have the capability to harness the intelligence talents of lots of countries in our behalf, not just the United States. You know, we have a very close working relationship with what we call the Five Eyes.

Peter Robinson: Right.

Amy Zegart: This is, you know, the

Peter Robinson: New Zealand, Australia, Canada, United Kingdom and us. Correct?

Amy Zegart: Exactly.

Peter Robinson: Oh, do I get extra credit for that?

Amy Zegart: You do. You get A plus.

Peter Robinson: Oh, excellent. All right.

Amy Zegart: And so we're seeing greater sharing among allies in Europe, too, not just the Five Eyes with Russia and Ukraine. So we have allies, the Chinese don't. They've got North Korea on their side. This is not a good team that you wanna be a part of from an intelligence perspective. We have values on our side. And I think if Ukraine has shown us anything, it's the power of those values. You know, I was doing research for this book on counterintelligence, I spent a lot of time looking at history at people who served United States and betrayed their own country, in this case, the Soviet Union in the Cold War. Why did they do it? And one of the most valuable assets we had was a guy whose code name was Top Hat. He he was a military officer in the Soviet Union. He never asked for a penny. He lost his life by giving secrets about the Soviet Union to the United States. Why did he do it? He believed in what we stood for. He believed in our freedom. And that values piece is enduring in intelligence. It's why foreigners will want to stand up, and risk their lives, and risk their families to help the United States. And I think the third, actually, I would say two more enduring values. We still have an innovation ecosystem that's second to none. So, and I think we can continue to use that. We can use it more effectively. But it hasn't died. We still have it.

Peter Robinson: All right.

Amy Zegart: And then we have this incredible multicultural society. We have people in this country that speak lots of different languages. They come from lots of different places. They understand lots of different cultures. What an intelligence advantage that is to understand governments and societies outside the United States, that the Chinese can't match in the same way that we can. So our immigrant background, our multicultural society, is an enormous intelligence advantage if we use it in the right way.

Peter Robinson: Let's close where we opened. One of your students here at Stanford comes to you for advice. She has a job offer from Google, interesting work, fascinating work, all kinds of technical challenges. Very good pay. Lots of her friends are going to be there. But she's also received expressions of interest from the CIA. Less pay. It's a government bureaucracy after all, when all is said and done. But the recruiter makes it sound surprisingly interesting. Professor Zegart, she says, what advice would you give me? How do I weigh these up?

Amy Zegart: Well, the first thing I always say to students is you have to follow your heart. That, and the advice is what it's worth, right? It's free. So they have to find what's right for them. But if it were me making that decision, I would say to myself, do you wanna sell ads and make algorithms work faster? Or do you wanna save lives and work for your country? I think the other thing too is, there are windows of opportunity where it's easier to go into government and it's easier to go into the private sector. And so I always tell my students think about what's a one-way street and a two-way street. Easier to go right from college into government service. And then you can leave and you can still serve your country in other ways, as you go to Google later, you go to the private sector later. And then by the way, you've taken that experience with you. You know, early in our conversation, we talked about why don't Silicon Valley leaders get it. Well, they haven't had that experience in government. And so government service or public service, I think we have to have a broader idea of that. It doesn't mean you're a lifer. You can be an ambassador from that experience working an intelligence agency, then go to Google and bring all that knowledge with you about how the government works, what that perspective is. You'll serve the company better. You'll serve your nation better. So I would say you don't have to choose, but if you had to go first, I'd say, go to the CIA first, and then pursue your private sector career, if you decide you want to. But I think the mission is such a strong pull, it'd be hard to leave.

Peter Robinson: Thank you. I'm a little sore about that one because all my life I've waited for the tap on the shoulder, and I've never been asked to join an intelligence agency. Amy Zegart, author of Spies, Lies, and Algorithms, the History and Future of American Intelligence. Thank you.

Amy Zegart: Thank you so much, Peter.

Peter Robinson: For Uncommon Knowledge, the Hoover Institution and Fox Nation, I'm Peter Robinson.

overlay image