A Catalogue of Journalistic Malfeasance

The reporting on Edward Snowden has been dreadful. Is there a way to make it better?

Two weeks ago, when the Guardian first leaked a Verizon court order to hand over its call metadata, a national debate began about privacy and security. Since then those leaks have continued, and they still drive the conversation. But much of that initial reporting turned out to be wrong — so much, in fact, that I’m starting to wonder if it’s approaching journalistic malfeasance.

The initial story, reported by Glenn Greenwald, seemed to expose a heretofore unknown vast surveillance apparatus operated by the federal government. “It is not known whether Verizon is the only cell-phone provider to be targeted with such an order,” the article said. Greenwald then wrote that “previous reporting” suggested other phone companies have received similar court orders.

While the story as reported sounds like normal journalistic practice, it is not. For one, there is still no evidence about how many other phone companies have been compelled to hand over their records. On Twitter, Greenwald wrote, “The program we exposed is the collection of all American’s [sic] phone records.” That isn’t true — he exposed the collection of Verizon’s records. The only evidence that this is an ongoing, long-standing program involving other telecos is a statement by Senate Intelligence Committee chairwoman Diane Feinstein and various anonymous leaks to national security reporters.

“NOBODY thought [the Patriot Act] enabled mass, indiscriminate, bulk collection of all Americans’ records,” Greenwald wrote on Twitter later. This statement, too, beggars belief: The assumption that the Patriot Act does underlies nearly all opposition to it, and has since 2001. Moreover, the metadata Verizon hands over to the NSA is legally collectible: Not only is it protected by a 1979 Supreme Court decision, but it has a clear law enforcement function as well.

The next leak Greenwald published, with veteran national security reporter Ewan MacAskill, made an even more eyepopping claim: “The National Security Agency has obtained direct access to the systems of Google, Facebook, Apple and other US internet giants.” That report also turned out to be largely exaggerated. Experienced tech journalists immediately seized on the description of the PRISM program (which accesses the data of internet companies) and poked it full of holes. Despite being described as “data-mining,” PRISM is really “nothing of the sort,” according to journalists who have covered the NSA in detail.

Rather than “direct access to the systems” of such internet companies, it turns out that the NSA has an automated system to request and serve subpoenas that companies then fulfill to their liking — through a dropbox, a secure FTP server, or some other method. It is not data-mining as many reports initially said. Soon after those first, misleading stories were published, the companies themselves rushed to issue public statements revealing how they cooperate and why, and how they try to fight back.

But the exaggerations and misreporting did not end there. The Guardian also published a twelve-minute-long interview with Edward Snowden, the source of the leaks and an NSA contractor who once worked for the CIA. Many of Snowden’s claims stretched credibility:

• He made considerably less money than he claimed. I assembled at least five other inconsistencies about Snowden’s story in this space last week.
• He claimed to have “been a spy most of his adult life,” when he worked in IT and ran a network. IT work is not spying, even if it’s classified.
• He claimed he could “access any CIA station in the world.” Robert Deitz, a former top lawyer at the NSA and CIA, told the L.A. Times that claim was a “complete and utter” falsehood. It’s unclear who is really telling the truth, but that’s important: there is no way to verify Snowden’s claims and they should not be printed as fact.
• Snowden said he participated in a CIA operation to “recruit” a Swiss banker in Geneva through a manufactured drunk driving arrest. Swiss President Ueli Maurer over the weekend said that such a claim “does not seem to me that it… played out as it has been described by Snowden and by the media.”
• During Congressional testimony, NSA chief General Keith Alexander called Snowden’s claim that he could tap into anyone’s email “false.” He continued, “I know of no way to do that.”

Indeed, a common thread in the flurry of reporting around the NSA revelations was a seeming misunderstanding of the technology involved, leaving reporters susceptible to dubious claims they did not know how to verify.

As a result, Snowden and the reporters following his every word have stretched basic facts. In addition to the details above, the program of mass surveillance on Americans has been discussed openly for years. In 2009, the Obama administration revealed that it had to scale back some activities that had become ethically suspect. Then, last week, a common, unclassified software package — PRISM — badly misreported. Journalists did not distinguish between a software program (like the web browser you’re using to read this post) and a program of operations. They printed that the software program was secret — it was not, as a five-minute Google search revealed — and missed that the operation using PRISM actually was. It was a total mess.

The Guardian isn’t the only paper to have been caught up in the sexiness of the Snowden leaks. The Washington Post also had to walk back much of its early reporting. Over the weekend, CNET got caught misreporting comments Rep. Jerry Nadler (D-NY) made about NSA surveillance. In their initial story, they reported he had said the NSA told him they could listen to phone calls without a warrant. In fact, Rep. Nadler said nothing of the sort and had to rush out a clarification to reporters.

If there’s any common thread holding all of these misreportings and poor fact checking together, it is time. The rush to be first out of the gate with explosive new details of anything — or, in the Guardian’s case, the rush to publish before Snowden could be located and arrested — created perverse incentives to publish without verification. Washington Post freelancer Barton Gellman even said that his attempts to verify some of Snowden’s claims led to Snowden pushing the same documents to the Guardian because they would publish faster.

Yet the rush to be first has also led to false or deeply misleading stories about the scope and purpose of government activities. Some of this misreporting goes viral, cementing untruths in the public that corrections after the fact will never repair.In the meantime, revelations that more than half the Senate skipped a classified NSA briefing last week go mostly unremarked upon, as do government claims that these NSA programs thwarted attacks in more than 20 countries. The U.S. is now openly concerned that Snowden could be an asset working for the Chinese government. Those stories aren’t nearly as sexy as the apparently shocking revelation that governments spy on each other.

This same problem also surfaced during the Boston bombing: News organizations rushed to be first with new information, and most of it was just wrong. Ultimately, that is just a cost of doing business in the breaking news game.

So what’s the solution? For one, stop assuming the first version of the “facts” is correct. So much of the initial round of NSA reporting has turned out to be false or misleading that it’s a wonder such misreporting hasn’t become its own scandal. The speed with which false information propagates in the public (and worse, in commentaries) is dismaying to those of us who’d prefer public debates be based in fact rather than fiction.

Yet so long as breaking news dominates the coverage, there will continue to be frenzied periods of rushed reporting and eventual retractions or clarifications. Until we as people change our media consumption habits, the news organizations that continue to rush poorly researched information into the public record will have no reason to change their ways.