Inside WikiLeaks

by Daniel Domscheit-Berg

  We decided to get the media involved right from the start but remain in control of decision making. So we looked around for reliable partners. We soon decided on the New York Times. For strategic reasons, we wanted an American publication on our side, and why not start with who we considered to be the most prominent and influential? Our second major partner was Britain’s Guardian newspaper, at which Julian claimed to have good contacts. In Germany, we decided to work together with Der Spiegel, the country’s most respected weekly newsmagazine. They were my responsibility.

  Marcel Rosenbach, Holger Stark, and John Goetz are three experienced journalists who work in Der Spiegel’s Berlin office. We had first really attracted their attention after the publication of the “Collateral Murder” video and they had contacted us for the first time at the Re:Publica 2010, a Berlin conference on the Internet 2.0. I provided them with a completely encrypted laptop so they could safely store the documents, and our media partners equipped them with Cryptophones. The irony was, we never spoke on the phone.

  We met once a week to keep one another up to speed and ensure that everything was going well. We had agreed on a date, July 26, 2010, on which both they and WL would simultaneously publish the material. That date was still several weeks away.

  The journalists went through the material and did their own original research. We made sure that the documents would be processed technically in preparation for the story as a whole to go online. But the first problem had already cropped up.

  We wanted to cooperate with a number of media outlets and let more than just those three publications in on what we had. But journalists are like dogs jealously guarding a bone when they think someone is trying to take a story away from them. The publications we already had onboard all wanted exclusive rights to stories.

  Marc Thörner, for example, was a German journalist who has written extensively and excellently about Afghanistan. He reported from the country for a long time, and his book The Afghanistan Code was widely praised by the press. We wanted to involve him in the background research and allow him, too, to view the documents. But the other publications turned up their noses. Some fly-by-night freelancer was being cut in on the deal? Large news publications could never allow something like that to happen. In their eyes, they played in a whole different league. Thörner, who would later write the most thorough report on the subject, would only be allowed to publish a day after the “big three.” Although we had begun by saying we would never give up our ultimate authority over the details of how and with whom we cooperated, the freelance journalist was pushed to the side.

  For me, Thörner was nonnegotiable, and I had said so to Der Spiegel, but the Guardian and the Times piled on additional pressure. Julian guaranteed them exclusivity. As confrontational as he often behaved toward us, he was completely tame with the journalists from those two newspapers. I know, of course, that it’s difficult to get your way with the media, and there was no doubt that these guys had more experience than we did. What had we been thinking? Getting exclusive, hot news was their stock in trade. We should have known that they would try to impose their rules on us.

  Our original plan was to get everyone together in London. There, they would lock themselves in a basement somewhere and confer about the material—similar to the procedure used for the “Collateral Murder” video. We had agreed not to reveal to the journalists that even more revelations were waiting in the wings. In addition to the 91,000 documents, some other equally controversial material had arrived. Time constraints had only allowed us to a take a general look at it. But we suspected that we were sitting on a powder keg.

  Nothing worked out as planned. Julian flew to London by himself, refusing offers of assistance from me and the others on our core team. I later heard that our colleague from the New York Times made it known immediately that he wanted to work in his own home office and took off back to the Big Apple as soon as he had copied the documents onto his laptop. Even documents about the Iraq War, which had never been part of the deal, ended up on his hard drive. Then he hopped a plane. That violated all the agreements we had made, but Julian permitted it.

  David Leigh from the Guardian took over the job of coordinating. The guys from Der Spiegel told me that Julian often appeared at meetings looking completely exhausted or immediately submerged himself in work on his computer. Soon, it became clear that we were no longer in control of the process. We were completely swamped by the task of technically processing the documents. Our technicians were working round the clock to put the documents into a readable format.

  The date of publication was set for a Monday so that Der Spiegel, as a weekly magazine, could stick to its normal schedule. In return, the magazine departed from some of its usual practices. Parliamentarians in Berlin would not receive their usual advance copy on Sunday, and the electronic version would be delayed until we had gone online.

  On the Wednesday before publication, I met Rosenbach and Goetz for lunch in an Italian restaurant in central Berlin. I wasn’t very hungry. The menu was far too selective for my tastes, but to be polite, I ordered a bowl of pasta. While the other two talked, I twirled the noodles around my fork. The two journalists said everything was going swimmingly. I watched the ball of spaghetti around my fork getting larger.

  “And you guys?” Goetz asked. “Everything going all right with you?”

  I took a bite and nodded. The Spiegel journalists looked happy. Goetz was rubbing his hands. I got a bad feeling. I completely lost my appetite, though, when they asked me how far we had gotten with the “harm-minimization” process.

  “Have you finished the blackouts?” he asked.

  I stared into space like an idiot, fighting to keep my facial expression under control. They had agreed with Julian, Rosenbach reminded me, that we would black out all the names from the documents. That was a condition all three publications had set. It was essential before the material could go online.

  I didn’t know anything about it. It sounded logical to delete the names of innocent parties, and I didn’t have any personal objections to it. But during this phase, Julian only ever told me crucial details very late in the game. That often put me in a difficult position with journalists. It was entirely possible that this was the explanation this time as well. I hurried home.

  I immediately contacted our technicians and their assistants. They were all drowning in work. It was the first any of them had ever heard about blacking out names. We were caught between a rock and a hard place. The reports were as good as done, and the printing presses were already warming up. It was too late to stop production. It would have cost Der Spiegel, as a magazine, untold thousands of euros to cancel a publication date planned so far in advance.

  I logged onto the chat room. Julian was there.

  “Hey, what’s all this about harm minimization?” I asked.

  In an instant, he was gone. He didn’t reappear for the rest of the day.

  For everyone else, things were coming to a head. We did everything within the realm of possibility. I think that from that Wednesday to the following Monday I got only twelve hours of sleep, if that. Anke was living with a ghost.

  A glance at the documents revealed that even when you blacked out all the names, enough contextual information remained to identify the people involved. If, for example, a report said that one of the three Afghans captured on March 25, 2009, in village X had turned informant, it would have been child’s play for the local Taliban to find this person and exact retribution.

  Ninety-one thousand documents! It was simply too much. I stared at my computer, at my wit’s end. There was no way we could do what we needed to with the raw material. We required an Internet interface that would make the editorial omissions easier. Our technicians would subsequently develop a program with which large numbers of volunteer assistants could securely access the documents and black out the names that occurred there. But for our imminent publication, any such help would arrive too late. There was nothing we could do.
<
br />   Thankfully, our media partners handed us a solution. They recommended that we sort out 14,000 of the documents and hold them back for the time being. The material in question was the so-called threat reports. They contained assessments of levels of risk based on warnings received from Afghans. If an Afghan knew about an insurgent attack and wanted to prevent it, for instance, he would get in touch with American troops. They would write up the information in a threat report, citing their source, making the informants easy prey for acts of Taliban retribution.

  In the remaining 77,000 documents, names cropped up only very sporadically. Various people from the media later checked the documents and found only around one hundred names. A Pentagon spokesperson nonetheless said WL had “blood on its hands.” It turned out that none of the informants were harmed by our publications. And it later emerged that the US Defense Department had concluded in an internal paper that the information we had leaked wasn’t dangerous.

  We were working as fast as we could go, when Julian turned up again in the chat room. He had been intending to tell us about blacking out names that evening, he wrote. He also presented us with a comprehensive to-do list:

  J: 1. the urls need to be standardized tomorrow. the naming has been standardized. “kabul war dairies” and “baghdad war dairies”

  J: 2. afg needs to be checked for innocent informer identification. These are mostly in the threat reports. its quite a bit of work to go through them

  J: 3. high level overview and press release need to be done

  J: 3.5 our own itnernal coms must be standardized. sat pagers deployed if available and silc/irc fallbacks

  J: 4. distribution infrastructure needs to be tested again

  J: 5. versions of the afg database that we supply need to have the classification field stripped out

  J: 6. i have made a full sql version of the database that also needs to be put up as one of the downloadable archives

  J: 7. torrents seeded / archives pre-deployed

  J: 8. email machines need to be made robust.

  J: 9. press team / contacts standardized

  J: that’s it for the things that MUST be done or we fail

  J: now for those things that need to be done if we are to do justice

  J: 10. i have the perl based searchable/explorable front end i and the guardian developed. that also needs to be deployed as a downloadable archive [more on that later]

  J: 11. a short 3 minute video intro needs to be made. I have people here ready for the film/editing part, but the graphics part (e.g. google earth / ground images) needs to be done

  J: 12. the people [journalists] who worked on the data all need to be interviewed about their approach and the qua ities/limitations of the data. 10 to 20 mins each. no prep is needed. i have this assigned at the london end, but we also need to do berlin and new york. this is a fast way of producing a “guidebook” for the material, and also elevates WL into a clear working-in-partnership with these three major players

  J: 13. the press team needs to be robustified and we need a list of talking heads to can speak sensibily about the issues (not just us)

  J: 14. donation systems need to be checked / and made slightly clearer / the australian po box needs to be put up for cheques etc and possibly the .au bank account should also be expoed

  I wrote back, putting into words what everyone was thinking: “It’s only 4 days until the release …”

  The hint about omitting the threat reports had come from our media partners. We hadn’t had any chance to familiarize ourselves with the content of the documents. That was the journalists’ job. Nonetheless, Julian would later get up in front of the cameras and rhapsodize about himself and his “harm-minimization” procedures.

  Our technicians, too, put in hundreds of hours of work without ever getting thanked for it. For example, they converted everything into KML format so that WikiLeaks users could locate every individual incident on a timeline at Google Earth. Even without Julian’s to-do list, the timing would have been incredibly tight.

  Naturally, the night before we were scheduled to publish, we still weren’t finished. The Guardian simply went online without us. The New York Times waited because they were afraid of being the only ones in the United States with the publication. And the people from Der Spiegel called me every hour, asking when we were finally going to put the leak on our site. It was utter chaos. But once the media machine had revved up, no one cared whether we had messed up the teamwork a bit and lagged behind our partners in publishing the material.

  Around the world, there was a huge debate about whether this publication caused anyone any harm. There was less discussion of the actual content, aside from the first wave of coverage about the details and a second wave of analysis, once experts had digested the material. Julian had said that it was his mission to end a war. If he’s succeeded, it’s news to me.

  We had expected the documents to change people’s thinking about the war. If everyone could see how much wrong was being done in Afghanistan, we thought, people would protest and demand that their governments bring home the troops.

  One problem was the sheer volume of data. The collection of material was too large for people to enter into the debate simply. And we still had 14,000 documents containing even more explosive stuff. Most of the reports published by Der Spiegel, the Guardian, and the Times were based on the latter. It was ultimately very much worth our three media partners’ while to get to exploit this material exclusively, while the competition had to make do with the leftovers.

  Naturally, it would be unfair to attack individual journalists for wanting exclusive access to good stories. I personally had good relationships with most of the reporters I knew. But the way the media functioned—the addiction to information possessed by no one else, the constant desire to squeeze as much as possible out of a story, the mix of permanent curiosity and friendly arrogance—disgusted me sometimes. I thought back to the days when we weren’t so well known. When I had to contact the media to solicit their interest in good material. When they didn’t call me back or answer my e-mails.

  The majority of journalists viewed us with a critical eye during our first year of operation, and some wrote clever analyses of the problems our platform entailed. That was fine by me. Some of them changed their tune, however, when they realized how much attention our material could generate. They began to suck up to us. I found that quite strange.

  Amid the debate engendered by the leaks around this time, criticism was growing ever louder that WL’s only enemy was the United States. There were many corners of the globe, critics argued, that equally deserved to have the spotlight turned on them. And in fact, all of our major publications in 2010 were aimed at the one remaining superpower, the United States.

  There were many reasons for this. First of all, Julian’s frustration with American foreign policy was fed by the simple fact that the United States played a leading role in most of the world’s major conflicts. Moreover, the suspicion could hardly be dismissed outright that the United States waged war partly for economic reasons. US intervention in the affairs of foreign countries was a particularly grievous sin. But one could have legitimately argued that it was equally important to criticize countries that committed crimes against their own people.

  Another, more banal, reason why we focused on the United States was the language barrier. None of us spoke Hebrew or Korean. It wasn’t easy to gauge the significance of a document even when it was written in English. Julian speaks no foreign languages at all. Even though he exploited his advantage as the only native English speaker among us in internal discussions and diverted uncomfortable debates into pedantic lessons about the precise meaning of this or that word, he himself was often unable to remember the names of non-English media partners or collaborators. In a TV interview after I had left WL, he even tied himself in knots trying to pronounce my real last name. We should have tried to find more volunteers to help us with translations, but we had already failed to integrate as
sistants for far more basic tasks.

  The third and most significant reason for our focus, though, was that by homing in on the United States, we were seeking out the biggest possible adversary. Julian Assange had no time to tussle with lightweights. He had to single out the most powerful nation on earth. Your own stature, it has been written, can be measured by that of your enemies. Why should he expend his fighting energy in Africa or Mongolia and get into quarrels with the Thai royal family? It would have been a far less attractive prospect to end up in some jail in Africa, or wearing concrete boots at the bottom of some Russian river, than to inform the world that he was being pursued by the CIA. And it wouldn’t have gotten him on the nightly news.

  The biggest problem we had in conjunction with the publication of the Afghan War Logs was that Julian had gotten ahead of himself and showed the journalists our additional material. That tied us to our existing partners. Our plan of remaining masters of our own destiny had become a farce.

  The New York Times, for instance, had cited WikiLeaks as the source of the leaked material, presumably out of concern that the link might bring them into conflict with the law. But they already had the material on Iraq. It would have been nearly impossible to stage the next link without them.

  Some weeks after the Afghan War Logs leak, the Washington Post ran a long story titled “Top Secret America.” It investigated the background of the military-industrial complex and convincingly illustrated how the sector had grown enormously as a result of the so-called War on Terror. The report was full of really good information. I don’t know where it came from, but it was a true journalistic achievement. The piece also included numerous graphics and applications generated by the Post’s own editorial staff. People there had asked me whether they could have access to the remaining 14,000 documents. That would have been a sensible instance of cooperation. I would have liked to have rewarded their fine work with a part of our secret treasure.