Arms and influence

Inspecting the CIA

The latest podcast from the National Spy Museum is especially good. Frederick Hitz, the first inspector general for the CIA, covers a number of critical topics, including weaknesses of the Director of National Intelligence (DNI), the critical importance of the staff for Congressional oversight committees, and the uphill battle hiring good people into intelligence jobs at painfully sub par salaries. Definitely worth the half hour listening to it.

People may give a brief nod to the weakness of the rational, unified actor model of government behavior in foreign policy and national security. Yeah, yeah, the government is really different formal organizations and distinct personalities that cooperate or clash. There's no philosopher king with complete control over everything. If you really want to understand why governments do what they do, you really have to understand how their constituent parts work together (or not).

Unfortunately, it takes real attention to learn how governments really work. Changes over time, like the upheavals in the CIA over the last couple of decades, only make it that much harder for the average citizen to keep track of it all. Yet another reason why we need fewer lazy journalists who attend press conferences at the Pentagon or the White House and type up the results. To return to the content of the Hitz interview for a moment, it isn't enough to report the passage of the Intelligence Reform and Terrorism Act. The real story is what happens next.

So much wrong, so little right

Last week's story about NSA employees eavesdropping on Americans abroad calling home to the United States was no shock. Things like "greatest hits" of phone sex and other intimate conversations, recorded and passed around NSA offices like joke e-mails, are just the sort of abuse you'd expect when you grant ordinary people extraordinary surveillance powers with no oversight.

As is typical of these sad counterterrorism tales from the Bush years, there are lots of bad sides--and no good side. NSA employees violated the privacy of innocent Americans in humiliating ways. The entire program exists under a legal and Constitutional cloud. And no one is claiming that this surveillance led to the capture or killing of any foreign terrorists, or the prevention of a single terrorist attack.

By now, Americans should have seen enough of these failures to accept the dodge, "We'd love to tell you about the successes, but we can't." We gave up the instruments of good government, designed for both difficult and easier times, so that we could fight Al Qaeda better. Instead, we got sweaty civil servants eavesdropping on husbands and wives in their most private moments.

Sometimes, the abyss stares back

After the 9/11 attacks, many people who had little or no experience in counterterrorism were casting about fearfully and desperately for ways to prevent another attack. We're aware of many of their mistakes, but this one certainly is in a category all by itself:

The military trainers who came to Guantánamo Bay in December 2002 based an entire interrogation class on a chart showing the effects of “coercive management techniques” for possible use on prisoners, including “sleep deprivation,” “prolonged constraint,” and “exposure.”

What the trainers did not say, and may not have known, was that their chart had been copied verbatim from a 1957 Air Force study of Chinese Communist techniques used during the Korean War to obtain confessions, many of them false, from American prisoners.

Two things make this incident particularly odious. First, of course, there's the idea that US officials would copy the interrogation techniques used by our enemies in the Cold War. However, you might give them the benefit of the doubt if the methods actually worked.

However, it's important to understand exactly what these techniques were designed to do. Controlling behavior, not eliciting accurate intelligence, was the goal. The Soviets had applied similar techniques long before the Cold War had started, in many cases to get public confessions from "counter-revolutionaries," who in most cases were just people unlucky enough to lose Stalin's trust, but gain his attention. Foreign nationals, such as the leaders of the WWII Polish underground (pictured here), and captured American pilot Francis Gary Powers, were also the subjects of show trials.

Public political points, not the quiet gathering of accurate intelligence, was the endgame for techniques like "Semi-Starvation," "Exploitation of Wounds," and "Filthy, Infested Surroundings." In fact, these methods made it harder to get reliable information from the victims, since they were so broken down that they were eager to tell their captors whatever they wanted to hear.

During the Cold War, anti-communists in the West pointed to show trials and brutal interrogations as evidence that the USSR and other communist regimes were an evil on par with Hitler. While we might dicker over who deserves the "Most Evil" title, Hitler or Stalin, it's enough to say that the anti-communists were correct. Books like Darkness At Noon, Animal Farm, and The Gulag Archipelago gained talismanic power because they described an evil that needed to be confronted. They were never written to be how-to manuals.

If you think that I'm expending a lot of words on what may have been just the horrendous bad judgment of one Guantanamo Bay "instructor," I'll just add this final observation that might make these words worthwhile. Excessive secrecy has not been protecting interrogation methods that work. Instead, they have been keeping the public eye from seeing exactly how unprepared, unprincipled, and ineffective some of the people on the "front lines" against Al Qaeda have been.

[P.S. For those who don't know me, such as someone who posted in the comments section of one of yesterday's posts, I make these arguments as a person who spent a good part of his adult life worried that Americans were not taking the threats of terrorist and guerrilla groups seriously enough. I also worried that Americans did not realize how dreadfully unprepared the US was to fight these enemies. Unfortunately, after 2001, a lot of people who knew next to nothing about these forms of warfare devised strategies that they argued were necessary and effective, but have proved exactly the opposite.]

Now hear this

Some excellent podcasts recently that deserve your attention:

• From This American Life, the most screwed-up terrorism prosecution imaginable. It's so screwed up that it's hard to figure out exactly what happened.
• Also from This American Life, a journalist discovers that he's the target of a warrantless wiretap.
• From Spycast, the International Spy Museum's podcast, a veteran counterterrorism professional talks about the nature of the terrorist threat and the self-imposed problems the United States has in fighting it.

There's a common theme among these three podcasts. Whoever the next President will be, whatever the face of the next Congress, the question people will be asking in 2009 is, "How on earth did we waste eight years combating terrorist groups? And how do we hold people accountable for screwing around?"

Tourist intelligence

If you're ever in the DC area, I strongly recommend visiting the International Spy Museum. Don't worry about the hokey "You're a spy now" introduction. There's a lot of more substantive stuff to come.

Be sure to look through the book section of the gift shop. I dare you to walk away without buying a book you didn't know existed and now need to have.

The FBI's NSL FUBAR

To no one's surprise, the FBI abused the expanded powers for domestic surveillance granted by the PATRIOT Act.

In a review focusing on FBI investigations in 2006, Justice Department Inspector General Glenn A. Fine found numerous privacy breaches by the bureau in its use of national security letters, or NSLs, which allowed the FBI to obtain personal information on tens of thousands of Americans and foreigners without approval from a judge.

Why NSLs are a BFD
If you need a refresher, NSLs gave the FBI the ability to poke around your personal information, such as finances, e-mail, and telephone calls, without the approval of a judge, grand jury, or even a prosecutor. Worse, if you are involved in this collection--for example, as an employee of a phone company turning over your cell phone records--you could not discuss the NSL with anyone. No asking your supervisor or lawyer whether this request was proper and legal, and most of all, no telling the target about the records search.

In 2007, a federal judge struck down the NSL portion of the PATRIOT Act. However, the decision dealt with the principle behind NSLs; it did not answer the question, How often did the FBI abuse this power?

The Inspector General's report expands the story beyond the less-than-credible internal FBI investigation:

According to Fine's report, the FBI continued to rely heavily on national security letters in counterterrorism, counterintelligence and cybercrime investigations, issuing nearly 50,000 of the documents in 2006 alone. Nearly 200,000 were issued from 2003 through 2006, the report said, and were used in a third of all FBI national security probes during that time.

Even more important than the scope of abuse is its cause:

The pattern persisted in 2006, Fine concluded in the report issued today, in part because the FBI had not yet halted the shoddy recordkeeping, poor oversight and other practices that contributed to the problems. He also said it was unclear whether reforms enacted by the Justice Department and FBI last year will address all the issues identified by his investigators.

So much for technology
It's worth digging into some details of that "shoddy recordkeeping." A few years ago, you probably skipped any articles about the FBI's problems implementing a "case management system." However, this story, which got no attention outside a few journals that cover computer technology in the federal government, is perhaps one of the best examples of how things went horribly wrong with counterterrorism during the Bush years.

Since most people haven't heard of a case management system before, here's a quick explanation of what it is. Many legal and government jobs are all about opening and closing cases. For example, a lawyer needs an efficient way to collect and organize the information about a particular court case. At the same time, that lawyer's boss is scrutinizing how quickly and effectively the lawyer handles the case, so the case management has an important managerial function as well. The case management system, for everyone from trial lawyers to FBI agents, is where that person spends a large, important amount of time each day. (If you want more information about what these applications do, click here for the American Bar Association's ratings for various case management systems.)

For several years, the FBI tried, and ultimately failed, to implement a case management system. The following headlines from Government Computer News give a nice summary of what happened:

• FBI plans to build new case management system from scratch (12/30/04)
• Senators fume as FBI admits Trilogy foul-ups (02/04/05)
• Cold case (03/07/05)
• FBI takes another swing at case management with Sentinel system (05/24/05)
• Report: FBI ‘scrambling’ to launch case file system (06/06/05)
• Justice, FBI to overhaul fingerprint and case management systems (08/29/05)

As the FBI's case management project crashed into a brick wall, the FBI kept on issuing NSLs at a furious rate--200,000 between 2003 and 2006, according to the Inspector General's report.

Last year, the FBI finally announced that it was ready to launch the new system, six years after the 9/11 attacks. The old case management system, based on 1970s-era technology, was already a failure, since many FBI employees avoided spending time entering data into it. (That's a familiar problem, by the way, with many systems that ask people to stop what they're doing and type up their notes.) All talk about "service-oriented architecture features that facilitate information exchange among law enforcement systems" aside, there's still an open question about getting people in the FBI to use any system, new or old.

If this were the only example of information technology (IT) projects in the FBI that went south, you might chalk it up to bad luck, peculiar difficulties with this sort of system, or the team working on that particular project. However, the FBI has fumbled many IT projects, including the useless terrorist watch list database. In 2007, another Justice Department IG report found that the FBI was losing laptops at a rate of 2.6 per month.

So, let's summarize:

• The Bush Administration pushed to expand the warrantless surveillance powers of the executive branch.
• The "point of the spear," the FBI, has depended on antiquated systems to store, secure, and analyze this information--when FBI employees were using these systems at all.
• A major overhaul of the case management system, needed to track sensitive information collected via NSLs and other mechanisms, was a long, expensive failure.
• There's no evidence that the White House put pressure on the Justice Department to fix these problems.

Should the results be a surprise?

There are certainly people to blame, such as the Chief Information Officer of the FBI, Zalmai Azmi. However, as implied in the bulleted list above, it's also up to the President and his staff to pay attention to these details. It's also important for the US public to mind these details more carefully. Millions of taxpayer dollars spent, thousands of breaches of privacy, and not one terrorist attack stopped.

"Endemic surveillance"

It's clear from this analysis that countries electronically monitor their own citizens because they want to, not because they can.

The top tier, evocatively named "endemic surveillance societies," include Russia, China, the United Kingdom, France, Singapore, Malaysia, Thailand, Singapore, and the United States. If surveillance were merely a matter of capability ("Hey, we have all this eavesdropping technology already manufactured or widely sold in this country--let's put it to use!"), you'd expect more European countries, and perhaps Canada, to appear on the list.

Instead, political considerations seem to be behind "endemic surveillance." The countries in the top tier fall somewhere on a continuum between concerns about terrorism (say, the UK) and an interest in maintaining internal political controls (for example, China). Surveillance at this level isn't free, so you'd expect motivation, not capacity, to be a key factor.

The good news is that terrorist attacks don't automatically inspire a high level of eavesdropping. Countries like Canada, Italy, Spain, and Germany have certainly had their fill of terrorism; however, they haven't jumped to the highest level of surveillance because of potential "enemies in our midst." They also have a pretty good record of fighting terrorism, so it would be wrong to conclude that it's only the counterterrorism milquetoasts who stay away from heavy surveillance. On the flip side, it'd be hard to say that Russia has an admirable counterterrorism record.)

In a way, this Privacy International study is good news. The science fiction writer David Brin argued that endemic surveillance is a technological inevitability, to the point where privacy will cease to exist. The facts appear to contradict Brin: many societies choose not to monitor their own citizens, even though they could purchase and deploy the necessary technology.

Among some of the happier conclusions from this study: "The privacy performance of older democracies in Europe is generally failing, while the performance of newer democracies is becoming generally stronger." In other words, the less jaded you are about democracy, the more likely you are to protect privacy rights. Something worth remembering during the campaign season here in the United States.

[Thanks to Lawyers, Guns, and Money for the pointer.]

Revenge of the Gosslings

One side of the missing interrogation tapes story that hasn't been discussed enough is the following question: To what extent is the culture of the "new CIA" to blame?

By the "new CIA," I mean the infusion of less experienced but politically orthodox employees that spiked when Porter Goss became Director. In 2005, Goss' appointment started a wave of resignations, including some very senior, very experienced people from the covert action side of the CIA. Goss and his entourage had been hostile to the traditional Agency for years; seeing them walking in the front door, many career CIA employees decided to exit through the back.

Goss may be gone, but his legacy remains. A decade or so ago, the people who worked at the CIA inherited generations of experience at spycraft. That legacy was often quite literal: some CIA operatives were the children of former operatives. The CIA's history was definitely marred by major errors, from the failure to predict the collapse of the USSR to the immoral testing of LSD on unwitting subjects. Still, the Agency learned some of the important lessons from its successes and failures; certainly, the decades of institutional experience were a vital commodity.

The Agency also learned the inevitability of legal discovery. From the Church Committee through the FBI probes in the 1990s, CIA employees have had ample reminders that their actions might fall under legal scrutiny.

However, that's probably not the expectation of the "Gosslings." Instead, many of them brought to the Agency the "pull no punches, be bound by no law" attitude that the Bush Administration encouraged throughout the executive branch.  In other words, some of the people who might have been involved in the decision to destroy the interrogation tapes were guided not by decades of hard-won CIA experience, but the fantasies of the Fox scriptwriters behind the 24 TV series.

It's therefore pointless for Porter Goss to throw up his hands and claim that he had nothing to do with the destruction of the tapes, any more than Henry II could be held blameless for wishing aloud that something would be done about that meddlesome priest.

Interrogation that works

This excellent op-ed piece by Stuart Herrington describes how interrogation really works:

In interrogation centers I ran, we called prisoners "guests" and extended military courtesies, such as saluting captured officers. We strove to undermine a prisoner's belief system, which we knew instructed him that Americans are unschooled infidels who would bully him and resort to intimidation, threats and brutality. Patience was essential. We rejected the view that interrogators could merely "take off the gloves" and that information would somehow magically flow if we brutalized our "guests." This notion was uninformed and counterproductive, not to mention illegal, and we made sure our chain of command understood that bowing to such tempting theories would result in bad information.

In other words, you want to "turn" someone, to increase the amount and reliability of information. You don't want to turn them into someone who will lie to you, just to get the excruciating pain to stop. If you doubt Herrington, first go read his book about his experience as a military intelligence professional during the Vietnam War (link to your right).

[Thanks to Misfire! Misfire! Misfire! for the pointer.]

The best spies money can buy?

Let's hope that we see more attention paid to the "outsourcing" of US intelligence gathering and analysis. Just as Americans were unpleasantly surprised to learn how many military tasks, such as piloting Predator drones, were handled by civilian contractors, they're likely to be just as flabbergasted that a rapidly increasing number of intelligence functions are now privatized.

Worse, the argument is often no more specific than, "The competition of the private sector is more likely to produce better results than a bunch of civil servants." If there were more of a "business case" to let businesses handle American spycraft, the $1 billion of privatized line items in the latest Defense Intelligence Agency budget might have a bit more specificity:

The DIA did not specify exactly what it wants the contractors to do but said it is seeking teams to fulfill "operational and mission requirements" that include intelligence "Gathering and Collection, Analysis, Utilization, and Strategy and Support." It holds out the possibility that five or more contractors may be hired and promised more details on Aug. 27.

If you want to see if your vague feelings of unease about intelligence privatization have any justification, read RJ Hillhouse's blog, The Spy Who Billed Me. Hillhouse, who has been a voice in the wilderness on this issue, details the risks, such as the possible slanting of the President's daily briefing. Judge for yourself how serious these risks are.

Privatization of government functions does not get the skeptical questioning it should, particularly when we're talking about fundamental national security responsibilities. Here's a slow softball pitch to the Congressional Democrats: demand that privatization be doubted, not automatically granted the benefit of the doubt.