Inside WikiLeaks

by Daniel Domscheit-Berg

  We would also have liked to set up some sort of feedback system, a channel to communicate directly back with our sources. Of course, one basic idea of WL, and a main security guarantee WikiLeaks offers to informants, was that there’s absolutely no way to find out who our sources were. On the other hand, it would have been helpful for journalists to be able to contact them. But that would have been going too far. If we had turned reporters loose on our sources, the latter wouldn’t have been able to protect themselves at all. Based on my experience, I wouldn’t advise any informant to contact the traditional media with a digital secret document, not even if that person had a personal contact or was offered a small financial reward for the material.

  There had been one case comparable to Manning’s in the past, although the material involved was not nearly as incendiary. It involved student fraternities. Fraternities had become something of a running joke for us since their secret ritual handbooks so regularly turned up on our submissions platform. We could have filled a whole set of shelves with books from fraternities. Kappa Sigma, Alpha Chi Sigma, Alpha Phi Alpha, Pi Kappa Alpha, Sigma Chi, Sigma Alpha Epsilon, Sigma Phi Epsilon—we had them all.

  Among other things, these books contain the initiation rites used to torture new members—occasionally to death—and the secret codes, symbols, and songs of these groups. The scenarios described in their pages ranged from altars containing a skull-and-crossbones and a Bible to flags that had to be hung to the right and left of windows to a chemists’ fraternity’s demands that its pledges bring a long list of substances, presumably to be stolen from the lab, to their initiation for a bit of perilous magic. At the bottom of that list was a fire extinguisher.

  We asked ourselves whether fraternities were important enough that their handbooks merited publishing. We decided in the end that every pledge had a right to know what he was letting himself in for, so we put all this nonsense online. And once we had published one handbook, we had to be consistent and post all the others as well.

  These leaks earned us great animosity. Members of Alpha Gamma Whatever were always popping up in our chat room, and soon we felt like we could identify a frat boy from the first sentences he wrote. The message usually began: “Really awesome.” Then a pause and: “I think it’s totally great what you’re doing.” And then there’d be a sentence like “I’ve got a question about a publication.…” Sometimes, we’d answer directly: “Say, are you a member of a fraternity?”

  That was fine until someone gave us a handbook that had been reproduced using a digital camera. These handbooks always have a number on the first page, telling you which university it’s from, and at every university there’s someone responsible for keeping the book secret. Our source had manipulated the photos to black out the telltale number, and we converted the files to PDFs and published them. But someone also uploaded the original photos onto a forum that was frequented by fraternity brothers. It wasn’t hard to make out the number from the back of the first page. That revealed which university the book—and thus the source—had come from.

  Fraternity members, upset at the betrayal of trust, began to scour pictures on university servers and college communities, comparing the metadata from them with the metadata from the photos of the handbook of rituals. As a result, they soon found out the owner of the camera and the presumed informant. Things could have gotten pretty bad for him.

  Fraternities copyright everything, from all their songs on down to the last idiotic secret gesture or handshake. But not their secret rituals. That was fortunate for the person they accused of being the leak. Because fraternities were paranoid about someone finding out about this secret, they didn’t even show their handbooks to a copyright office.

  Our loyal chat guests were deeply unhappy that we were exposing their secret associations to public scrutiny. Once they’d realized that we weren’t going to remove the books from our site, they reacted with rage and, more frequently, whining. Occasionally I used to discuss the matter with them. They argued that nothing was more important in their lives than their fraternity. Paternal advice from me, to the effect that they should wait ten years and they’d feel differently, did nothing to alleviate their sense of loss. Once their secret rituals and codes were on the Internet, they could no longer be sure who was and wasn’t one of their fellow brothers. One of the main reasons for secrets is people’s desire to share them only with a select circle and exclude everybody else. Fraternities are a very vivid example of this.

  If it were true that a person like Bradley Manning was the one who uploaded the material we used to produce “Collateral Murder,” I could understand why he had done so. Manning was in his midtwenties, a person sitting in Iraq, isolated from his normal social contacts and probably surrounded by soldiers who had completely different attitudes toward war from his. The person in question may well have felt the need to talk to someone about his experiences, and that could have been why he acquired the video material and documents.

  It was probably impossible to expect people to keep knowledge like this to themselves. Probably most of our sources contacted us because they felt they simply had to share with others what they knew.

  Working for WL taught me that secrets are almost never kept. I wonder whether there even is such a thing as a secret between two people. I think they are very, very rare. If a sentence began with the words “I’ll only tell you if you promise not to pass it on,” it was nearly a foregone conclusion that this promise would get broken in another sentence beginning with those same words. This sort of prelude only prevents a secret from being spread quickly; despite what people may have promised, the secret would still make the rounds in the end. Even if someone’s best friend or spouse were the one being told, the danger of revelation was there—at the latest when the two parties in question got into a fight.

  Whoever copied the “Collateral Murder” and diplomatic cables material was running a huge risk, the dimensions of which may not have been clear to the whistle-blower himself. Perhaps he suspected that he was doing something forbidden, but he would not have known the extent of potential consequences. More than likely, the person in question was driven by moral considerations. But regardless of whom we have to thank for the material, any source should be told repeatedly not to talk about it with anyone. It would have been good if one of us had been able to offer this counsel to our sources.

  However, the anonymity guaranteed by WikiLeaks’s anonymizing mechanisms is the main advantage WL enjoys over all classical forms of investigative journalism. In most countries in the world, no journalist can guarantee a source, in any serious way, that his name is safe from the pressure and legal instruments of investigative authorities. The technical and legal construction of WikiLeaks, on the other hand, ensures whistle-blowers’ anonymity.

  But legal security is only one part of the story. In the course of our work, we repeatedly saw how naively most journalists handle technology and modern means of communication. Sensitive documents are anything but secure on the computer of the average reporter.

  When was a document so dangerous that we could no longer publish it? We discussed the topic a lot, particularly in conjunction with the leaked diplomatic dispatches. After Manning’s arrest, we posed the question somewhat differently: When was a document so dangerous for the source that we should no longer publish it?

  Theoretically, the problem can arise with every publication. What were we supposed to do, for instance, when a source contacted us three days later and asked us to delete his documents? Should the source always have the final say?

  That was precisely the question we faced once with a leak concerning Italy—one that ironically proved to be of little interest to anyone. The leak was about the questionable way in which a contract had been handed out, and the source contacted us a few days after we had published it and requested that we withdraw our allegations of corruption. I replaced the word “corruption” with a milder formulation in our summary of the document’s content, but I didn’t remove
the document itself from the site. That wouldn’t be easy, from a technical standpoint, if it is indeed even possible.

  The result was a series of questions. How could we be sure that the source wasn’t making a request for withdrawal because of external pressure? How could we be certain that, if we gave in, others sources wouldn’t be put under similar pressure in the future? And how could we know if it even was the actual source who was making the request? We concluded it would be in everyone’s interest to institute an ironclad policy of automatic publication after submission. As soon as anyone decided to upload documents on our site, he had decided they would be published. We had to draw the line somewhere.

  In return, we constantly tried to devise ways of preventing innocent parties from being adversely affected. We aimed to consider every aspect that could be problematic for the real people named in the documents or for the source. Those aspects varied from case to case. For example, it meant that we sometimes deleted names—or at least addresses and telephone numbers. We weren’t always as successful as we would have liked, though, and that was to become the biggest problem we would face with our Afghan war diary and Iraq war log leaks.

  Nonetheless, it was important to send a signal that exerting pressure on sources was not going to yield the desired result. That no matter what happened, we were going to publish the material we received. I think, by and large, this was a sensible decision.

  We had received the “Collateral Murder” material and diplomatic cables from someone—whoever that might have been—and we had already published the video. Manning was then arrested. Given the opaque nature of the situation, we should have ruled out any further publication of the American documents. Every new release risked providing further impetus for investigations of anyone suspected, rightly or wrongly, of being a source. I opposed further publication right from the start.

  There are lots of myths surrounding what led to Manning’s arrest. On the surface, the story was simple: He had chatted with Adrian Lamo, and that was what set the investigations in motion. Still, anecdotes and conspiracy theories flourished. There were some indications from the United States that the “discoveries” made by American authorities might not have been as accidental as they initially appeared. At Defcon—a security conference for computer technology, held in Las Vegas—in 2010, there was a lecture about a government project called “Vigilant.” According to the talk, security officials were involved in scanning the Internet broadly in search of relationships and data transfers that would reveal connections between people. If a lot of material was transferred between A and B, alarm bells would ring, and authorities would begin investigating. It’s entirely possible that people who worked for the US Army snooped around on their own servers. That wasn’t particularly problematic. Two million people in the United States alone had access to documents with the same level of secrecy as the cables. Intelligence services only got actively involved when material was obviously being passed outside that circle. And this, according to the story told at the conference, was how Manning had attracted their attention. Later, the US government would deny the whole strange tale.

  But there are also other, even more obscure, theories concerning personal motivations. Lamo himself claims he turned Manning over because he recognized the global political importance of the material and felt morally compelled to act. Ultimately, the question is to what extent a chat can even serve as evidence. It’s hard to identify people in chat rooms.

  It’s possible, though, that the whole story was much more banal. US authorities may have only decided, in retrospect, to present Lamo’s accidental discovery as the intended result of their own investigations. That, in any case, would have been a logical move.

  We will probably never know the whole truth. Proceedings in front of military tribunals aren’t public, and those concerned will invest considerable energy to ensure no one smuggles any information out of these tribunals.

  When people who obviously wanted to offer us material appeared in the WL chat room, they were usually directed to me. It was important to caution them not to reveal too much about themselves. We had a standard warning we constantly repeated: no names and no identifying information. We had to prevent people from writing things that would allow conclusions to be drawn about who they were. Our internal standards were quite high. After all, we subjected ourselves to the same restrictions on talking about material.

  Julian had a fine nose for especially interesting material and how it could be used to exert political influence. Discerning what was truly interesting was something we had learned over the years, from negative experience as well as positive. There would occasionally be documents we believed would be of interest but that failed to create a stir.

  One such example was a series of field manuals we received, including US Army handbooks on waging unconventional warfare. They described the methods used to weaken other countries from within and replace a head of state with a military regime. At the time, I thought this handbook would cause outrage around the world, and I expected journalists to beat down our doors. But no one cared, because the subject matter was too complex.

  Videos were another thing entirely. Even if the images only related to one concrete incident, we soon realized how great an impact a video could have. Julian, in particular, developed a keen eye for this.

  Julian was a hacker and the book he collaborated on, Underground, was highly regarded in the hacking scene. As much as he professed to despise hackers because they weren’t politically motivated, he was not a complete stranger to the heroically clandestine nature of what they did. But at its core, WL functioned completely differently. WL was a platform, a tool, an instance of technology devoted to absolute neutrality, not intervention.

  Julian’s later accusations that I was typical “middle management” likely give an insight into his way of thinking. Despite the fact that we were equipped with Cryptophones and worked with the curtains drawn, and although Julian mentally transformed innocent plane passengers into State Department spies, we were all just administrators, managers, and press spokespeople—anything but true warriors of the digital underground. We were people who rented servers. We waited for material instead of soliciting, contracting, or hacking it ourselves. That was the way we understood our jobs, and regardless of whether Julian found it sexy, it was important for us to be like that.

  It has been suggested that by posting our “Most Wanted” list WL was, in effect, soliciting material from its sources. This practice was patterned after the one from the Center for Democracy and Technology, on our site. It was intended to appeal to our potential sources’ sense of adventure, but it also pushed the boundaries of intervention. It must be said, however, that we hadn’t drawn up the list ourselves but rather asked our readers to make suggestions.

  Externally, we made it known that we, of course, would do everything in our power to support Manning without implying in any way that he had anything to do with the leaks. Julian announced that he was going to organize top lawyers and unleash a giant wave of media interest. He called for donations, saying $100,000 was needed to provide Manning with the best legal help. I arranged the server where the Support Manning campaign would run. Someone else was supposed to be responsible for the content.

  But early on, the initiative was already stalling. When I asked Julian about contacting Manning’s attorneys, I received no concrete answers. Journalists were constantly calling me, and their questions were persistent. The Association of German Scientists had gotten in touch, with the idea of awarding Manning their whistle-blower prize.

  I inquired about this with Julian, but his answer was:

  J: i have no time to explain that and given you don’t need to know it; next …

  J: i know why you were asking which makes it all the more infuriating

  D: so why am i asking?

  J: some moronic disinformation campaign

  D: no. i am asking because i am putting my ass out there on the line for an official position
that you have claimed, and that i get asked about

  J: lawyers names can’t be given. they’re not our lawyers names to give. They’re bradley’s lawyers, blah lbah

  J: you don’t need to know because you can’t tell people, bah blah, hence waste of time

  I have to admit that we at WL, myself included, utterly failed on this score. Unfortunately, as was too often the case, I simply yielded to whatever Julian said. I complained often enough about Julian being a dictator who decided everything on his own and withheld information from me, and I wouldn’t take back that criticism. But that didn’t absolve me of responsibility. I shouldn’t have let myself be worn down by the stress. I should have asked more questions and taken the initiative myself, if necessary. There was no reason why Julian alone should have been responsible for us supporting Manning.

  In the end, we took part in the campaign, organized by family and friends, that ran on the Bradley Manning Support Network. Later, we argued about how much we should actually contribute. Julian thought the $100,000 we’d asked our supporters to donate for his cause was too high a sum, and revised it downward to $50,000. As of this writing, $15,100 has been transferred from the Wau Holland Foundation to the Bradley Manning Support Network.

  WE had tried out a number of variations with leaks. We had simply loaded them onto the site without any fanfare; we had gotten individual journalists onboard; and we had held press conferences as a media organization. But in the summer of 2010, we were determined to do everything right. We were sitting on a huge pile of documents concerning the war in Afghanistan: 91,000 of them, to be precise, from the US Central Command. The Afghan War Diaries, as they were known, covered everything from status reports to information about firefights and air raids, details about suspicious incidents and so-called threat reports, and the procedures followed by American troops in combating insurgents. No newspaper, book, or documentary film had ever contained such profound and comprehensive firsthand insights into the war. It was simply fantastic.