How to Run a CIA Base in Afghanistan

"I spent a lot of time breaking the laws in other countries"

This month, Statecraft focuses on America’s attempts to build nations in Iraq and Afghanistan. We think these stories are a great opportunity to get a closer look at what we mean by “state capacity.” Massive stateside bureaucracies attempted to build new, stable institutions in foreign countries. How did those bureaucracies perform? How about the institutions we built abroad?

Last week, we talked to the former USAID Deputy Director of Procurement in Afghanistan, about how money gets spent and misspent in foreign aid.

How the Government Loses a Road

In today’s installment, we talk to Laura Thomas, a former Central Intelligence Agency (CIA) case officer and Chief of Base in Afghanistan. She has served over 17 years in national security and leadership roles. Laura was circumspect in our conversation, as you’ll see. That said, she shared more than I expected.

What You’ll Learn

• How a CIA station operates

• What kind of intelligence failure was October 7th?

• The holes in how CIA teaches tradecraft

• Why did intelligence agencies predict Kabul would fall quickly to the Taliban?

---

You had to go through CIA Publications Review Board (PRB) Approval for this interview. What would happen if you didn’t go through that process, or if we go outside of it and blow it here? 

Imprisonment for me would be one, if I shared classified information. That's definitely something I'd like to avoid. I spent a lot of time breaking the laws in other countries. I'd really like to stay in the good graces of my own. 

The PRB doesn’t change content and isn’t trying to moderate criticism. They’re trying to ensure that classified information that would threaten current sources and tradecraft. aka “methods,” aren’t out there in the public. Because it could lead to the arrest, imprisonment, and death of CIA sources, which would be really blowing it. That’s why there’s a review process.

Great. Let's jump into it. I would love an org chart. You were chief of base: walk me through the hierarchy. 

A base is a smaller subset of a station, usually in a different geographic area, but there's some nuance there.

So a chief of base is the person who's in a specific field location making decisions and authorizing any CIA operations that are going on in that location. They're also generally the person that is the senior most interlocutor with foreign partners in that location. Other case officers and individuals in a base would report up through the chief.

The chief also has release authority for cable traffic. Similar to the State Department, CIA sends cables back and forth between stations and to CIA headquarters. It's how information is processed, relayed, transmitted, how operations are approved, not approved. And release authority means you can hit the button and say this cable or intelligence report is going out into the wild, it will forever be memorialized. Someone thought it was a good idea and you have to stand behind what you put out there. 

There are other roles in a station: targeting officers, collecting officers. Talk to me about those positions. 

One thing that's shifted over the last decade is just more people going out to the field than what was traditional.

When you see the movies or you read, especially older books, it's all about the case officer. The case officer, or C/O, is the person that's out recruiting the foreign sources, aka the assets, what people usually refer to as spies, and that certainly still exists. That's still the forefront. 

The main thrust of what CIA does is to collect human intelligence, also known as HUMINT, from these foreign sources, and then analyze all-source intelligence, meaning make judgements on it based on information coming in from a wide array of sources. And the case officer is the person who goes out, targets those sources, says, hey, this is a person I think would have valuable information for the US government, but also would be willing at some point to share that information in a secret, long-term relationship. 

The case officer’s day job usually consists of writing cables about HUMINT operations and writing up intelligence reports from assets that a C/O has recently met. Additionally, many days include meeting with other government stakeholders (State, Ambassador, Defense, Commerce, etc), as well as with foreign partners who are aware of our CIA affiliation.

There’s a lot of writing involved. Much time is spent on the specific wording of intelligence reports and cables to ensure accuracy and context on a source’s motivations and level of access. For cables, that’s what I call “verbal jiu jitsu,” because we have to navigate other Stations’ operations, the government of the country we’re in, other foreign governments in that country or interested in what we’re doing, other US agencies who believe they should be consulted in advance or at least looped into the details of what we’re doing, and so forth.

We’re constantly balancing all those equities with the need to move quickly. We have to move at the pace of events in the world. We’re constantly judging whether we have the right balance of action, thoughtfulness, and legal authorities, all weighed by the results we achieve or don’t achieve.

That’s just the day job.

At night (though really at any time) a case officer is out hunting for targets, for people to recruit who can provide sensitive information to the US Government and meeting those people in various scenarios and finding ways to advance the relationship. Or they’re out clandestinely meeting assets and debriefing them for intelligence that the C/O then has to write up and send into headquarters for further dissemination to the White House, Intelligence Community, and other USG entities. 

Other people in a base or station: A desk officer is a person that is essentially helping with the flow of approvals. You need operational approvals in order to recruit people, in order to do certain things. There are certain levels of authority that anyone in the US government has: operating authorities, you can authorize up to so much money to be paid out to this source or for this operation.

But a lot of times you have to go back and ask headquarters for approval. Desk officers help with that flow, making sure that everything is happening on time with proper approvals in a base and headquarters. They do a lot of basic research as well, just to make sure that the job gets done.

There are collection management officers, also known as CMOs. They're the people who interact the most with Congress and other policymakers. They find out what the U.S. government needs to know, and make sure that the case officers are asking the right questions of sources in order to answer those questions.

They're also helping edit intelligence reports for source protection before those reports are disseminated to the wider policymaking community. We really don't want a leak of an intelligence report if it's so obvious that it must have come from this specific source, in this specific ministry, in this specific country. That’s a failure on many fronts because it’s jeopardizing a source’s life. 

That would be catastrophic from a CIA standpoint, because the sacred obligation of CIA is to protect its sources. I didn’t look at them like, “oh, they’re just fodder, we'll recruit this person, we'll take their information, and who cares what happens to them.” There's a huge ethical issue with that, but also a practical one, in that if CIA were to do that, who would ever want to volunteer to work as a clandestine source? 

Targeting officers are the officers at CIA who basically write the book on a specific target. They are analyzing all sorts of information coming in, whether it's signals intelligence (SIGINT), HUMINT, open source, and they're creating a profile of an individual or perhaps a terrorist group that CIA wants to go out and recruit a source from or within, and really helps the case officer think about how they approach an individual and perhaps where to find that individual overseas.

They're extremely valuable. Great targeters are like gold. 

To clarify, is that targeting officer ever personally interacting with the target? 

No. It's not to say that it would never happen. I've definitely had targeters come with me to meetings. It's pretty neat for them to eventually meet a target that they followed for so long, but it usually does not happen. 

Why are more folks in the field in recent years? What's the dynamic there? 

I believe it's because of the war zones: Iraq and Afghanistan, just how many people were deployed to support those efforts. In a war zone, where immediacy is very important, as soon as you have a piece of information you're trying to pass it to the military for them to act on it. The more people who can help in a support function for doing that, the better.

This is a very controversial thing for many case officers like me: if you're winding back the number of case officers, to make room for other support officers in the field, are you limiting the amount of information that you can collect with the core human mission that you have? And that's certainly debatable. 

As you describe it, the whole team plays a role, but you're limited in what you can do by the number of case officers you have in the field. Is that a fair assessment? 

Yeah. But I think we’re more limited by bureaucratic risk aversion to human intelligence (HUMINT) operations on the whole. One thing that is so incredibly obvious, especially post October 7th: HUMINT has never been more important. SIGINT is great, SIGINT can be wonderful, but it only gets you so far.

You really have to recruit a human that can tell you not what's happening in the moment, but what is going to happen, because they intuit the plans and intentions of a group.

And you have to have case officers to go out and recruit those people. You can't just sit back and hope that SIGINT solves the problem for you. That's the easy button. The easy button doesn't always work. 

When you mention October 7th, do you mean that there was a relative over-reliance on SIGINT there? 

Yeah, absolutely. 100%. I think October 7th is an absolute intelligence failure on many levels. It showed an over-reliance on SIGINT, becoming enamored with technology and forgetting the basics, getting out and talking to people face to face. 

It likely also shows a reliance on foreign partners, and assuming that they have something covered and that we need to step back and just let them do their job. 

Is force protection a reason intelligence agencies fall in love with technology, or risk aversion with human intelligence? Does signals intelligence feel safer, or less risky? 

Yes, CIA is supposed to be the “collector of last resort.” That's the moniker that the intelligence community uses. And it means you only recruit a human source to provide information if you cannot get it by any other means, and we don't want to risk the life of a source, or the case officer, if it's something that we could collect via open source, SIGINT, or diplomatic reporting channels. 

You get addicted to SIGINT and technology. I think that's a major challenge. There's a mixture of over-reliance on technology, reliance on SIGINT, but also just hubris. Thinking you know a target so well, someone couldn't possibly be planning something that you've not anticipated.

This is not unique to October 7th, unfortunately, and this is human nature. We all do this in some aspect in our lives.

Cultivating human intelligence, that tradecraft: how does that set of skills get passed down?

There are training courses specifically for people handling sources and tradecraft training and methodologies. Some are tried and true and meant to withstand the test of time. So to speak with whatever technology advances occur at the same time. 

CIA is facing a real reckoning. It's called “ubiquitous technical surveillance.” It's essentially that there's some digital record of everything you do nowadays. 

CIA is facing a huge challenge in how it operates because of technology. But there are ways to double down and use technology to advance operations as well. And obviously I'm not going to be able to get into that. This is a big ongoing test of the agency. 

What challenge does that new paradigm pose to traditional ways of collecting human intelligence? Does it change how you identify targets? 

Operations have to be much more thoughtful, it slows you down to a significant degree. I think that it reinforces the fact that you have to be incredibly judicious in choosing operations – who you choose to recruit and how you choose to do it.

It should mean fewer human sources, only the most well placed sources, because you're running a risk every time you're running a human being. They're putting their lives on the line and it's not just the source's life, it's their family. Usually these people have spouses and children.

Let me go back to your time in Afghanistan. Where was your base? 

Can't say that. 

Okay. 

Yeah, it was a small base in Afghanistan. 

You mentioned the relationship between bases and stations. Will you walk through what that jargon means in Afghanistan specifically? 

Sure. So there were multiple, a number of bases at any given time in Afghanistan when the U.S. was there. Base covers a specific area: it could be a province, could be bigger than a province. It all just depends on the situation. And then the base reports up through the main station, which in that case would have been Kabul Station. 

And is the station always nation-specific? Do you ever have stations that are smaller or larger in scope than one? 

Yeah, so I can't get into too many details on where things might diverge from the typical, but some bases are incredibly large and some of the most senior people in the agency, for example, would be a base chief at some of these locations. 

Also, there are some stations that are incredibly small that don't have bases within them, and it might be a more junior person that's the station chief in that location. It really just depends on the size, where it's located, what types of operations are happening in that location. You may have base chiefs who are more senior to station chiefs somewhere else. 

Is that formal, on-paper seniority, or implicit/informal? 

 It's both. It's formal on paper. And it would be obvious to anyone who's worked at the agency for longer than a few months. 

When do you have to ask for authorization from station or from hierarchy back home? 

And how many of the directives on what information to gather come from back home, vs from developments in the field? Talk to me about the priorities. 

It all depends on the situation and it also can be very personality-dependent.

It depends on who is at headquarters running that specific group that covers your area, depends on who the chief is in the station or the base chief. Generally, if it's something really big and important, that's going to get the attention of the president or could somehow throw off relations with a very important foreign partner, headquarters is going to be involved. You want them to know if something like that is occurring. But it's all personality-dependent, except for things like financial authorities. That's pretty baked into any government institution. 

How much does the Department of Defense (DOD) make specific requests for intelligence? 

In a war zone, you're working with the military hand in hand, it’s a very close relationship. If the military had come and knocked on my door and said, “Laura, we've got a patrol going out today, here's where we're going. Can you tell us, do you have anything? Is there anyone you can ask?” Yeah, I'm going to start contacting sources immediately. I want to help prevent an ambush and loss of life, obviously. 

That sort of thing is fairly routine in a war zone. When it's a little more formalized, if you're giving briefings in D.C. or at the more senior levels to agency heads, or individuals at agencies that run major portfolios, you get their feedback. When they read intelligence reports, they comment on them, and then a collection management officer would take that feedback, and they would generate a list of requirements to ask sources. 

And they would choose which sources could answer these questions the best and which are worth tasking with these questions, because we have to be so sensitive about all aspects of our operations. It's a constant risk calculus. 

Is there a role or a kind of officer that you're typically interfacing with in that day-to-day role in a war zone?

Whomever they have nominated as the liaison. There's obviously a huge intelligence component of DOD, a number of different types of intelligence officers there. It’s usually an intel to intel liaison. 

But it would also depend on the location you're in and how small it is. If it's something where I can pick up a rock and throw it from one end of the base to the other, and the military's co-located there, I’m going to know all the military folks there and we’ll all talk. It may not be the formal mechanism that one would find in a much larger facility. 

Is there a specific win during your time on base that you can talk about more freely?

Not necessarily a win that I can talk about while I was at CIA, but one that happened just after I left that is very similar. I did a lot of work on exfiltrations, which we call “exfils” for short. This is when you help pull a source out of their home location because of a developing dangerous situation that you believe is going to result in their arrest and imprisonment or worse.

And the ones that I worked on, post-agency, was during the Afghanistan withdrawal in 2021, helping get a number of families out who were U.S. citizens or visa holders or had family members that had worked for US agencies or the military. It's the same exact skill set, similar processes. I certainly didn't have any authority, being outside the government at that point, but picking up the phone, vetting people who are requesting exfil, did they in fact work for the military like they said they did? What sorts of questions do I ask to try and vet their story? 

In some cases there's a language barrier, getting through that language barrier and being able to communicate with them, helping direct them around Taliban checkpoints, which gates to be able to access the Kabul airport compound, because that was a huge problem whenever the withdrawal happened. It was just complete and utter chaos. There were a number of gates, people were trying to flood the compound.

There was a huge concern: what if they flooded the compound to the point that it fell? Then you've got an even larger human catastrophe on your hands. Everyone probably remembers the visual of the military aircraft taking off and people hanging onto the landing gear.

On that topic, I know there was a lot of reporting contemporaneously on debates within the intelligence community about when the Afghan government would fall, when Kabul would fall. Tell me about those varying assessments. 

I didn't work there at the time. So I don't really know the accuracy or not. But it was no surprise to any mid-level person who served in Afghanistan that it fell so quickly, based on working with foreign partners there. For many of my colleagues who had also left, no surprise was registered with us at the speed at which things collapsed. It's unfortunate. This was a policy failure as well. We did not plan well and did not have ample time to get people out.

This should have been well planned and implemented far in advance. And it's debatable whether we should have left a small force there to hold the country, to prevent the takeover to begin with. 

When you say it was not a surprise, what kinds of things are factoring into that expectation? What made you think this was quite likely? 

I served in Afghanistan a long time ago, so this is not knowledge necessarily from my CIA days, but just having served there previously, following the news, having friends and colleagues in Afghanistan. The level of corruption in the country, the hubris, to think that being in a location and helping prop up a government and trying to stabilize it is going to be enough to hold it against a force like the Taliban. 

That's what I mean when I say it was no surprise that the Afghan government fell. It was always shaky at best. You don't change essentially an entire history and culture of people overnight. Twenty years is short when you think about global cultural shifts and the rise and fall of certain empires. You can't change it completely. You can try from the outside, but it also has to be something from within that takes hold, too. After twenty years, it had not fully taken root. 

So then what do you do? Do you sit, you say, “Oh I'm just going to stay in a country for a hundred years or more? And I'm the person who’s going to bring democracy?” Again, it goes back to hubris and being realistic about what is and isn’t feasible. 

Remind me the period in which you were with the CIA in Afghanistan? 

Can't say exactly, but well over a decade ago. 

At one point during our involvement, Afghanistan was the most corrupt country in the world, and more than 70 percent of the country was illiterate. Is there another way things could have gone here? Or was failure baked into the cake from the start?

I think this is a perennial challenge of US policy in general. A lot of times we get involved and we have one idea of what we think the outcome will be and it morphs into something completely different. And sometimes the best policy is just hands off. At the same time, we're very much about human rights and advancing them while trying to balance realpolitik interests.

I can't say what we could have done to make the outcome better. I do think that had we had a small force to stay in place there, it's possible it would have prevented the Taliban takeover, that they wouldn't have had the momentum they had, that the Afghan president would not have fled. I do think that we see some of the ramifications now of the withdrawal: Russia's invasion of Ukraine. When you get involved with a cause or take an action, you have to already start thinking about how you’re going to extricate yourself from it, and what would the second and third-order effects of real or perceived abandonment look like. 

Anyone who tells you they can predict the future, I think that's nearly charlatan-like behavior. It doesn't mean you shouldn't try and craft the future to some degree. I'm not saying that intervention is never the right idea, but I'm saying it's often the wrong idea. 

In our interviews this month, a lot of folks echo what you're saying about a talent gap. Agencies always seem to want more expertise and insight, more foreign language speakers, more experts. 

Is that a recurring problem in your experience? Just not enough folks with the kind of expertise you'd like to have? 

I think in general, if you don't speak a language and you don't live among the people that you're trying to recruit, or that you're making policy recommendations about, it is very hard to truly understand the psyche of people who come from vastly different backgrounds and experiences. 

Can you say more about that last point, about our own biases? Were there things that you saw on the ground that were more difficult for folks who were not there to process and make sense of? 

Yeah, in the US, we have this legacy and culture of thinking we're incredibly rational people. We're incredibly well educated in terms of lots of college degrees and equate that with being smart. We think reason rules, that people always choose what’s in their best interest, and forget the role of emotion, and basic human nature. We have a sense that's very palpable in the U.S. and West – and we can debate where it comes from – I think there's a lot of religious sort of undertones to it – but it is a very real sense of the value of human rights, the value of democratic principles. You're stepping into a world where that has not been the history of that country and people, it's a completely different trajectory. We often see the world as we want it to be, not as it really is. 

And there is thinking that within a year or two, you're going to import these values, because it's rational. We think, “Why would you not want to be free? Why would you not want a stable government system that provides human rights for all and judicial impartiality to all?” It just seems so obvious to people in the West, so we can’t understand when someone doesn’t want what we want.

You talked about training courses for human intelligence, and how that skillset and tradecraft is passed down. How granular is that training to country or region? 

It starts very broad and then as you go, it becomes more niche, and certainly once you start going into language training and in specific portfolio areas, there are additional courses that delve deeper into specific area considerations for tradecraft, but also for working with individuals from that specific culture. As much as expertise is important, there are some things that are just universal. Knowing how to listen, being polite, asking good questions, having a generally friendly demeanor. 

With surges of personnel into war zones, unfortunately not everyone has a background in the region. A lot of people don't speak the language. You're often working through interpreters.

You're trying to do the best job you can given the situation you're in and the lack of knowledge that you have. And that's where judgment comes in. And just trying to be as aware as possible of what you don't know and being careful about assuming that you know something that you don't.

And then leveraging others. There are a number of Afghan-American officers that work at the agency in other places and leveraging knowledge that they would have, or people who served in Afghanistan much earlier in the eighties and nineties and trying to understand their perspective. But you have to know their biases, too. They also represent snapshots in time or certain strains of thought that aren’t universally shared. CIA officers are always trying to sort through fact vs fiction and delusion, and like everyone, have to battle against their own self-delusion.

Are there particular cultural challenges in Afghanistan that are difficult for CIA officers to get up to speed on?

Yes, the vast history of the country and the breakdown of tribal dynamics. Not every officer, I think, is poring over every historical book and reading everything they possibly could before they deploy. They should be, but there's just not the time. That's a real challenge. Being able to talk about the history of a person’s country, tribe, or religion with them is very valuable to building a relationship. When I could cite an obscure event that happened in their country in 1920 and ask them how they think it affected the course of their own history, there is often a real appreciation from them that I’ve taken the time to try to understand.

When you have more time, you can step back and do that deeper dive. But again, I think there's a death of expertise that's happening. And I think the surges and frenetic pace of war zones have added to this.

And this is a big challenge for any government agency. Being a generalist can be great. And sometimes it's good to take someone who's worked on one area, and move them over to another because there's a lateral thinking that goes with that.

“And now this cross-collaboration and sharing is going to make us all stronger in some cases.” Sure. That's helpful. But in other cases, you're losing a level of expertise. 

When you learn a language and you're living with people, it’s very hard not to authentically care and want to understand their way of seeing the world. Some people think case officers are only transactional: “I'm going to recruit this person, I'm going to write up an intelligence report, I'm going to give them some money, and then we're going to call it a day, and I'm going to advance in my career.” But there's usually a real personal bond and relationship that forms, and the best case officers are incredibly authentic, and authentically want to know someone else from a true curiosity standpoint, not just, “How can I leverage this for the US’ gain and for my own career gain?” 

Obviously, we wouldn't be doing the job unless we were trying to help the US first and foremost. There has to be that north star. But there also has to be that element of actual personal care and concern if you're going to be successful over the long term. If I thought a certain case officer didn’t really care about this, then I wouldn’t want them in my base, and I would limit what operations and assets I give to them vs others who do.

What kinds of training are missing in foreign policy? Do folks need more history? Political science? Literature?

You're saying that ideally, everyone would have this in-depth sense of the history of the country they're going into. Are there ways those kinds of trainings could be better implemented at CIA? 

Yeah, study abroad. Getting people even just paid to live and learn a language overseas. Nothing beats being in a location that you're supposed to be developing expertise on.

And having seen it, there's certainly the ivory tower. Book learning is great. I think everyone should endeavor to read and truly take in a culture, a history, religion, the people. But books, seminars, and thinkpieces will only get you so far.

It takes real human interaction to truly develop expertise. Again, you're always making trade-offs because you're trying to move quickly. You're trying to get the job done. You can't afford to give everyone this ideal training. 

This is a point of frustration at times between CIA officers on the ground and those in the policy establishment in D.C., who maybe have studied a country at the graduate and PhD level, worked on a campaign, and worked in a think tank, but have never been “on the ground” for any meaningful period of time. There wasn’t much more frustrating than a call with someone on the National Security Council (NSC) who had jumped from campaign to think tank and back, who believed they were the foremost expert on a topic. 

CIA does not make policy, it only reports the facts on the ground and makes assessments. But a number of times in my career, it was a real challenge to watch people in D.C. get things so wrong and not be able to do anything about it. 

We are facing a dearth of people who truly have knowledge on topics nowadays, because we're surging people to different areas. We're trying to cross-pollinate and that's to our detriment most of the time. 

The biggest challenge, however, is mediocrity in bureaucracy.

That all goes back to how do we restore trust in government? If anyone can give you a concise answer to that, I think they're probably lying because there's no one way. This is a multivariable problem, but I think the first step would be how do you rid a bureaucracy of mediocrity? 

And I think that all goes to, what's your reward structure? How do you incentivize people to move up the ranks? How do you hold back people who aren't really contributing? Agencies have to reform themselves in that way – or you can have Congress do it, but Congress is completely dysfunctional too these days, so that's not going to work either. 

The CIA needs fresh thinking at the mid and senior levels. Growing risk aversion is natural as one moves up through the ranks in any bureaucracy. Granted, judgment can also grow as a function of experience, but not always. The challenge is to find and promote officers at the mid and senior levels who can combine both smart risk taking and judgment. Put those officers in pivotal decision-making areas in the field and back at the headquarters desk that approves risky but well-considered operations. 

The way to address this challenge at CIA is to implement an incentive structure that pushes quality of source/agent recruitments and operations involved in over quantity. Quantity can lead to grave mistakes. It pushes officers to ignore troubling aspects of a case that might suggest the source is a double agent because to write up those aspects or acknowledge them might mean a case officer doesn’t get his or her “recruitment.” 

At CIA, the level of quality and the recruitment of human assets should be downright surgical. CIA does not need to run a lot of human sources to have outsized and exceptional insight. It takes an incredibly long time to establish such networks and the growing world of ubiquitous technical surveillance (UTS) makes it exceptionally hard. But it’s doable, and there are strong officers who can navigate these challenges.

Addressing this challenge will help limit intelligence failures, but it won’t stop 100% of them. No intelligence service can, especially one that operates with democratic principles. CIA always gets the blame and rarely gets the credit – its mistakes are visible and its successes are not, mostly by design.