Saturday 27 November 2010

WikiLeaks

Rumours are coming out that WikiLeaks is going to publish yet more documents in the near future - allegedly they comprise of two million diplomatic wire messages between US embassies and the State Department.

The Government is putting a DA-notice on the data, meaning that British publications will not be able to publish details of the leak. It will be difficult to see how they think this will work as the WikiLeaks data will be available throughout the world. As the Spycatcher debacle showed, DA-notices are increasingly inapplicable in the modern world, where servers can be placed anywhere.

This presents many problems. Governments will have an obvious wish to keep embarrassing data secret, and whistle blowers have an obvious role in getting the truth out - the MP's expenses being a classic case to point. However, such data is also often complex to understand, and is often open to misinterpretation by non-experts.

But WikiLeaks is going far too far. Their previous large releases - 'Collateral murder', 'Iraq War Logs' and 'Afghan War Diary' are widely believed to have included information that put both soldiers and civilians in danger. WikiLeaks claim to have redacted all such data from the release, but that is widely disbelieved. Many think that their parsing was done in chunks - they put similar documents together, inspected a few, and if they saw things they did not want releasing, then they did not release the entire batch. Otherwise, the documents were released. The problems with this are obvious.

Why batch the data. From Wikipedia:

According to statements by Assange in 2010, submitted documents are vetted by a group of five reviewers, with expertise in different fields such as language or programming, who also investigate the background of the leaker if his or her identity is known. In that group, Assange has the final decision about the assessment of a document.

The 'Iraq War Logs' comprised 400,000 documents. It would take years for just five people to go through this data. Say you can parse one document a minute: that means, with a ten-hour day, the five reviewers can do 3,000 documents a day. Given these very liberal constraints, it would have taken them over100 days to check and redact all the documents. In reality, it would have been much more.


Therefore WikiLeaks rely on what they call 'volunteer journalists'. Note the word 'volunteer': the more people you have, the more you require rules, procedures, checks and balances to ensure that all the volunteers are working to the same standard. It also begs the question about whether the volunteers are qualified to make such judgements. The release of names of soldiers along with Iraqi and Afghan collaborators led Amnesty and others to make official complaints to WikiLeaks.


The amount of data is massive. The Telegraph studied the MP expenses data and analysed it, looking for the juicy stories and the best way of releasing the information to keep the story going (and their circulation up). They parsed the data and published the things that were, in their opinion, in the public interest. These made and shaped the story, and they did not do too bad a job of it.

Secondly, this data is all about the US. WikiLeaks has published interesting data from other countries (for instance Peruvian oil and Scientology), but this year they have been concentrating on the US. It would be nice if they published similar data from (say) North Korea, Russia, China, Burma, etc. It is increasingly looking like an organisation that is solely concentrating on the US state.

These new documents allegedly number 2,000,000 documents. There is no way that they have adequately parsed these to check that it is wise or moral for them to be released. And that is exceptionally worrying.

Of course, the US has an obvious problem with keeping their data secret, but that does not excuse WikiLeaks' behaviour. Even one WikiLeaks' founders is deeply critical of the organisation.

WikiLeaks could be a force for good in the world. Instead, I fear it is becoming a force of evil.

No comments: