Julian Assange, the founder of WikiLeaks, doesn't know who leaked the thousands of Afghanistan war documents that his site posted this week. That's not unusual—it's how WikiLeaks works. To get a scoop to WikiLeaks, a would-be whistle-blower clicks the Submit Documents button on the site's home page, then uploads a file through a form that encrypts every interaction between the source and the site. WikiLeaks keeps no logs of the submission, and the site says that it is legally bound, under Sweden's press secrecy laws, never to cooperate with any investigation into the identity of the source. The site takes several additional measures to scrub submitted documents of any information that could compromise the leaker, removing any ID trails left by word processing software, for instance. The site also constantly feeds fake submissions through its network in order to fool potential attackers. "We have never lost a source,"Assange declares in his pitch to whistle-blowers around the world. "None of our sources has been exposed or come to harm."
At the same time, WikiLeaks says its founding mission is radical transparency. Assange argues that "increased scrutiny"—of governments, corporations, and institutions like the Church of Scientology—can be a powerful force for good, reducing corruption and oppression. "Principled leaking has changed the course of history for the better; it can alter the course of history in the present; it can lead us to a better future," WikiLeaks says.
This is the paradox of WikiLeaks' methods. Is radical transparency compatible with total anonymity? If we don't know who the leaker is, why he's leaking, and how he came upon his information, can we really know the full story the document tells? More importantly, how can we know that the information is authentic? Look deeply into WikiLeaks' efforts at radical transparency and you find complete opacity; WikiLeaks wants to shine a light on the world, but only by keeping itself shrouded in secrecy.
Consider the Afghanistan war logs. WikiLeaks says that its source has given the site more than 91,000 classified military reports. So far, WikiLeaks has posted all but 15,000, which have been delayed "as part of a harm minimization process demanded by our source." What does this mean? You can imagine many different scenarios: Perhaps the source is affiliated with the U.S. military and is afraid that immediate publication of the unpublished reports might harm American troops. On the other hand, what if the "harm" that the source wants to minimize isn't physical, but political? For instance, perhaps the information is so incendiary that it would affect the war funding bill being debated in Congress or significantly impact the midterm elections. Or maybe the danger is entirely more intimate for the source. If the source is Bradley Manning, the Army intelligence analyst previously charged with sending documents to WikiLeaks, perhaps the demand to keep 15,000 documents in the dark is an effort to minimize further legal jeopardy.
Any one of these theories could be true, or none of them. That's the problem; the fact that the leaker wants to minimize harm suggests that he, like most whistle-blowers, has some sort of agenda. That agenda is a part of the story, and it could provide valuable context for all of this data. If the source wants the United States to end the war in Afghanistan, we would look at the documents in one way; if he simply wants the U.S. to prosecute the war differently—perhaps, for instance, by sending more troops—we would see the documents in an entirely new context. Because we know nothing about the source, we don't know the story he's trying to tell in releasing these—and not other—documents covering the war.
Also, isn't it at least conceivable that WikiLeaks' source altered these documents? Nobody in government has questioned the authenticity of the trove—in fact, in condemning their publication, the Obama administration seems to be confirming that the reports are real. Nobody in the media seems to be questioning whether WikiLeaks' stash is authentic either. This is partly because the documents look authentic—they correspond to known times and places of U.S. troops, and comment on real participants in the war in ways that fit a pattern with well-known events. But that only tells us that the documents aren't complete forgeries. There are tens of thousands of reports here, each extremely dense and technical. How do we know that the source didn't add, hide, or in some other way change a number or a sentence here and there in order to paint either a more positive or negative picture of the war?
We don't know. On its site, WikiLeaks explains its authentication process this way:
WikiLeaks staff examine all documents and label any suspicions of inauthenticity based on a forensic analysis of the document, means, motive and opportunity, cost of forgery, what the authoring organization claims and so on. We have become world leaders in this and have an enviable record: as far as can be determined, we have yet to make a mistake. This does not mean we will never make a mistake, but so far, our method is working and we have a reputation to protect.
I don't doubt that the site takes great pains to investigate its leaks—Raffi Khatchadourian's recent New Yorker profile of Assange painted a picture of tireless devotion. You might also argue that questions of trust are just part of the deal when you rely on anonymous sources—a long-defended, if controversial, journalistic practice. But there is a profound difference between how WikiLeaks uses anonymous sources and how the rest of the media does. When the New York Times has a document provided by an anonymous source, its reporter knows the identity of that source. In that case, we expect the reporter to assess both the source's information and the source's reasons for reporting it. When mainstream media outlets are duped by these anonymous sources, we—justifiably—blame them for not checking things out.
Indeed, some critics of anonymous sourcing have called for the outing of lying leakers. In October 2001, for instance, ABC News' investigative unit reported that "four well-placed and separate sources" revealed that the anthrax used in the postal terror attacks contained a chemical additive that suggested ties to Iraq. When, in 2008, the FBI revealed that its lead suspect was Bruce Ivins, a scientist with no connection to Iraq, Salon's Glenn Greenwald and journalism professor Jay Rosen both called on ABC to reveal its sources. Outing the sources, they argued, would serve as an important check on people who try to peddle false information—if a leaker knew that he could be outed for lying, he'd be much more cautious. (ABC News did not comply with the request.)
Any such checks on a source's veracity are impossible under WikiLeaks' processes. If WikiLeaks doesn't know who provided a document in question, how can it know the source's "means, motive, and opportunity" to determine if the document is real? Even Assange admits these shortcomings. He told reporters on Tuesday that his worst fear was that a source could subtly alter leaked documents without WikiLeaks finding out. If WikiLeaks does end up publishing a dubious document, there would be no way to expose its anonymous source in order to deter further fraud.
WikiLeaks says there is one additional check on the authenticity of its leaks—the "collective wisdom" afforded by the Internet minimizes the risks of fraud, the site says. If a document is fake, the "broader community" will suss it out. To me, that provides little comfort. WikiLeaks is not really a wiki—a Web site that allows the Internet hordes to edit its information. Although WikiLeaks shares what it calls the "comfortable presentation style of Wikipedia," it does not allow people to edit source documents nor the accompanying commentary that explain the documents (those are written by WikiLeaks staff, who aren't identified on the site). Also, unlike Wikipedia, WikiLeaks isn't an open organization that anyone can join, nor does it have any democratically agreed-upon rules about the kinds of leaks it will publish. Instead, in nearly every way that matters, WikiLeaks is an opaque, insular organization.
I'm not saying that any of this should disqualify WikiLeaks' documents from serious discussion. I am surprised, though, that the site's admirers—including Rosen and Greenwald—aren't more troubled by the secrecy at the heart of its operations. I also wonder whether the extreme measures of secrecy WikiLeaks grants to its sources are really necessary to its mission. If WikiLeaks really is legally bound to never release the identity of its leakers, can't it still find out, for itself, the identity of its sources—if only as a way to more thoroughly check out their leaks? Perhaps this would put WikiLeaks in some kind of legal or physical jeopardy; maybe Assange or his staff could be jailed or harassed for not identifying their sources. According to Assange, though, that's already the case—he says he could be jailed if he enters the United States, and he takes extreme security measures. These risks he takes to protect his sources are probably one of the main reasons the leakers trust WikiLeaks with their information. Given the site's track record, would many leakers balk if WikiLeaks began asking them simple questions? Let me offer a few suggestions: Who are you, how did you find this document, and why are you leaking it now?