The Guardian, which has its headquarters in London, cited "pressure from the UK government" as a reason for bringing in US partners to work on documents leaked by whistleblower Edward Snowden.
Last night James Ball from the Guardian and Jeff Larson from ProPublica, a Pulitzer-winning nonprofit, shared some of the behind-the-scenes story with those at MozFest, the Mozilla Festival, held in London this weekend.
They discussed how the three teams of journalists worked in collaboration to spot stories in the documents, with each team then writing an article in the style appropriate for their news outlet.
In June the Guardian first reported on on the Prism data mining programme by the NSA after Snowden handed files to Guardian journalist Glenn Greenwald. This was followed by a report in August which said the UK Government Communications Headquarters, known as GCHQ, had received funding from the NSA.
Later that month, on 18 August, Greenwald's partner David Miranda was held at Heathrow airport. Days later, the Guardian, which launched a US digital operation in 2011, revealed how and why it destroyed hard drives containing copies of the NSA files.
A few days later a collaboration with the New York Times was announced, and then ProPublica joined the team.
On 5 September, after a month-long investigation, the three outlets published articles all of which centred on the agencies' alleged ability to decipher "much of the online encryption relied upon by hundreds of millions of people", as reported by the Guardian.
Q&A: Inside the NSA crypto story
The question and answer conversation with James Ball and Jeff Larson was chaired by Dan Sinker, who heads up the Knight-Mozilla OpenNews project.
There were several elements that could not be discussed due to the nature of the investigation. Ball, who also worked for WikiLeaks, has spoken previously about secure communication and source protection relating to the NSA files.
Q. Why did the three organisations collaborate?
The investigation was "technically challenging" and resource intensive, Ball explained, requiring a full-time effort for four or five weeks.
And bringing in US partners also "helped with First Amendment issues", explaining how having three outlets based in two jurisdictions took advantage of press freedom laws. Indeed the three outlets said there was pressure from intelligence officials not to publish but editors decided the story was in the public interest.
"Though this is a difficult story – both to read and create – there are huge implications," said Ball.
Q: How much data needed to be reviewed?
Finding the story within the documents was a huge challenge, explained Ball. Where one WikiLeaks document would provide a front-page story, the nature of these files required decoding.
“It’s not a case of read a document and you get a scoop,” Ball added. "There maybe a one-line fact in the middle of a long document about something else."
"These are the crown jewels", Ball said, and closely guarded by the NSA.
The journalists therefore needed to understand how the agency classifies information, and get to grips with its "relationships with big business".
And it required background knowledge on cryptography. Larson explained how he had to talk to lots of experts to find out long it would take to break encryption, and how he "played around with it in his free time" and "read lots of books".
"You learn the art of asking hypothetical questions," Ball said, but highlighted the dangers of doing so and incorrectly implying wrongdoing.
Q: Why should the public decide what intelligence agencies can snoop on?
The public can only make an informed choice about what might be acceptable levels of surveillance if the extent is known and following a debate, Ball said.
In the US “there’s a debate brewing”, Larson explained, and Ball pointed out that a three-hour discussion is due to take place in Parliament this week. “That would not have happened without this. And that’s in the UK where the response has been minimal,” he said.
He explained that in the US a bill, which has a good chance of being passed, is going to be introduced into Congress with "some serious NSA reforms". "Almost anywhere in the world except the UK it’s being discussed as a legitimate debate," he added.
Q. How did the collaboration work on a practical level?
The collaboration "went pretty smoothly", Larson said. Following "a few days of meetings" they "sat down with the data and started going through it". Individual journalists would then spot stories.
"We shared all the information," Ball said, and discussed findings with one another to check "we thought it was about the same thing".
The three teams each wrote a story in the news outlet's style based on the facts uncovered.
Q: How do you avoid getting paranoid?
Larson explained the "nightmares" that happen as a result of working on such a sensitive investigation. "Code words are flying at you," he said.
Ball said it "could drive you mad" so instead "you have to stop worrying and just get on with the task".
Q. What are the implications if news organisations on tight budgets cannot afford to report such stories?
Investigations such as this are expensive, with bills for lawyers and travel, Ball, who has been travelling back and forth between New York and London, explained.
"Without well-resourced newsrooms you couldn’t do it," he said.
Further notes from the session are at this link.
Free daily newsletter
- The Week in Good News, a weekly newsletter from the NYT, acts as an antidote to the daily news cycle
- NYT uses new interactive design to report on sexual consent on campuses
- Shifting focus from offers to promoting its journalism, The New York Times continues to build its subscriber base
- What types of projects suit collaboration? Here are 9 examples to get you started
- Tip: Use these free training materials from ProPublica's Data Institute