Today, WikiLeaks has started releasing some 5 million internal e-mails (obtained through Anonymous) of Stratfor, a “subscription-based provider of geopolitical analysis” known for massively selling its services to corporate actors, various branches of (mainly US) government and beyond. According to their website, “Stratfor uses a unique, intelligence-based approach to gathering information via rigorous open-source monitoring and a global network of human sources”.
In its leaked 2007 ‘Glossary of Useful, Baffling and Strange Intelligence Terms’, Stratfor defines ‘Open Source’ as:
USG definition: everything in the world that we haven’t classified secret and above. Real definition: stuff that’s on the internet. Sometimes cheaper than humint. Frequently much more expensive and less reliable.
Assuming that the above source is genuine (although it can hardly be validated externally yet), both definitions are obviously funny given Stratfor’s own ‘approach’, but what is worse, both are plain nonsense.
Regarding the alleged US government definition, information that has been classified above ‘Unclassified’ by an appropriate authority, and that has not been unclassified through a FOIA request etc., cannot reasonably be called ‘open’. Even if it became publicly available for whatever reason (e.g. a leak), it would still not be legally accessible for anyone not having a clearance and the respective need-to-know.
But even if we’d stick to the author’s ‘real’, that is pseudo-realistic definition, it hardly makes more sense. Firstly, most of the information Stratfor compiles into new analytical products stems from open sources. Secondly, and as one example out of many others, one HUMINT source handled as ‘Geronimo’ (now come on!) is paid 1,200 USD a month for information of dubios quality – or for reading his local newspapers on Stratfor’s behalf. As compared to a Stratfor enterprise license for 20,000 USD a year, a subscription of the New York Times comes to 300 USD a year. So without questioning the general value and necessity of human sources (and journalists), I wonder if Geronimo and Stratfor’s other informants are an actual bargain, especially as HUMINT is not opposed to OSINT, but must be considered as an integral element of the OSINT enterprise (and opposed to TECHINT, as we define it). If for Stratfor only the the internet is an open source, then they are either blind or simply ignorant.
Still, in another document called ‘Security Information and Instructions’ (2009), Stratfor gives advice to new employees in order to uphold the company’s highfalutin self-perception as “the most credible, truthful and definitive global intelligence organization in the world”. It requires staff to “practice stringent security in accordance with industry standards”. In an e-mail from 2004, the company’s CEO, George Friedman, even went so far as to claim: “Everyone in Langley knows that we do things they have never been able to do with a small fraction of their resources.”
And, as you can imagine, he does not tire of sharing his cloak-and-dagger wisdom on open sources in yet another 2009 e-mail: “And that is the weakness of the government vision and why I never use the term open source in relation to us. […] The government confuses the dissemination of information with the collection of information. We are not at all like the nyt because their reporters always identify themselves as such. We may or may not as it suits us. This is the fundamental difference between journalism and intelligence. The source may never know he was a source, which is the definition of clandestine.” Has Mr. Friedman ever heard of investigative journalism before?
At the end of the day, it seems like the inner workings of Stratfor, such as “get[ting] some of that ‘leak-focused’ gravy train”, have themselves become little more than “stuff that’s on the internet”. That stuff’s worth remains to be seen.