How WikiLeaks Decrypted and Released the Video

A leaked file on its own does very little. Plenty of documents have been dumped online and vanished without trace. What happened between the arrival of this recording and the press conference in Washington is the reason anybody remembers it.

What arrived

The material came in through WikiLeaks' anonymous submission system, encrypted, as part of a much larger disclosure from a US Army intelligence analyst that eventually included hundreds of thousands of documents.

An encrypted file is not a story. Someone had to unlock it, work out what it showed, establish that it was genuine, and decide what to do with it.

The months of work

In early 2010 a small group assembled in Iceland to work on the file. Iceland was chosen partly for practical reasons and partly for legal ones: at the time the country was debating some of the strongest media-protection legislation in Europe.

The work divided into a few tasks.

• Decryption and technical handling, so the recording could be watched at all.
• Verification. Establishing the date, the location, the units involved and that the file was what it appeared to be.
• Research. Reading the existing public record, including news reports from July 2007 and a book published in 2009 that described the same engagement.
• Field work. Sending people to Baghdad.

Going to Baghdad

This is the part that separated the release from an anonymous upload. WikiLeaks sent researchers to Iraq to find the district, locate the families of the men who had been killed, and confirm that the event in the footage matched what people there remembered.

That work established details that appear nowhere in the military record, including what had happened to the children in the van and who the driver was. A great deal of what is known in English about Saleh Matasher Tomal exists because someone went and asked.

What going there actually established

Details that appear nowhere in the military record: what happened to the children in the van, and who the driver was. Much of what is known in English about Saleh Matasher Tomal exists because somebody went and asked.

The decision to make two versions

They produced the full recording of roughly thirty-nine minutes, and a shorter version of about seventeen minutes with captions, arrows identifying the people in the frame, and an opening quotation.

Both were published at the same time. The short one was what nearly everyone watched, and the differences between them became their own argument almost immediately. That argument has its own page.

The title

The name Collateral Murder was chosen deliberately, as a claim. It is the single most consequential editorial decision in the whole release.

It gave the material enormous reach, because it was an argument rather than a description, and it handed critics an obvious line of attack: that a publisher who had already reached a verdict could not be trusted with the evidence. Both effects were real and both were predictable, and the organisation chose reach.

Working while expecting to be stopped

The group operated on the assumption that they were being watched and that the release might be prevented. Copies were distributed, arrangements were made for the material to appear even if the people preparing it could not, and the timing was kept tight.

Whether that caution was proportionate is debatable. What is clear is that the release went ahead on the day planned, in a room full of reporters, in the capital of the country whose military had recorded it. That morning is covered here.

What WikiLeaks bought with all that preparation

Credibility, mostly. The authenticity of the footage was never seriously disputed, including by the US government, which is remarkable for a leak of this kind and is a direct result of the verification work.

It also bought specificity. The release did not present an anonymous atrocity. It presented a date, a district, named victims, an identified aircraft callsign and a documented investigation. That is why the story could not be waved away as unverifiable, and it is why, sixteen years later, an archive like this one can be built on public records rather than on rumour. Those records are listed here.

What they published is described in the main account, and the reason no lawful route had produced it is set out in the Reuters requests.