There was some criticism expressed on Wednesday about my contention that even if social media conversations were to be passed by internet service providers (ISPs) then this would create such a mass of “chatter” (the technical term for unimportant bulk communications data) that the services would be faced with an almost impossible task in pinpointing the key intelligence.
“Aha!”, said the critics, “in the case of the murder of Lee Rigby, the message was sent en clair. There was no encryption. It must be easy for the ISPs to set up a programme looking for key words such as “murder” or “jihad” and this will quickly highlight any dangerous information.” Sadly this is not so – and don’t just take my word for it. My old boss, Richard Barratt, former head of counter-terrorism at MI6, who also served with MI5, was questioned about this very issue on Radio 4’s Today programme yesterday. He said: “Facebook has about one and a third billion users and about five billion posts a day, so clearly on a worldwide basis it would be almost imposible to deal with the amount referred.“
You see, the problem is, and always has been since before 9/11, not that the intelligence services do not have enough information , but that they have too much. This was the key conclusion of the Congressional Commitee of Inquiry into 9/11. The trick is not to get more information, but to get the right information. A desk officer only has so much time to assess the intelligence in front of him (or her).
Seeing as I am forbidden from discussing my intelligence work here, an illustration from my law enforcement days might assist. As is well known, the electronic communications of specific criminals are often bugged to determine when their next robbery/drugs smuggling operation/kidnapping will take place. But even where we were tapping specific people for specific periods, it was extremely difficult to turn that raw data into specific leads. This is for a number of reasons: criminals disguise what they are talking about; they use family members to pass seemingly innocuous messages; often they seem certain to do something on a particular date, but then for some reason they fail to get their act together and suddenly do the deed on a completely different date (remember the key conversation from Adebelajo in the Rigby murder was sent a full six months before the murder was actually committed). Imagine how much more difficult that task is when you are asking untrained organisations to scan billions of messages and then to select the important ones. You might as well ask them to send you the phonebook.
An objective observer might sympathise with the unnamed Silicon Valley executive quoted in the Guardian who said: “Given all the information they have, with and without our permission, it is outrageous that they should try and blame Facebook. The conclusion of the [ISC] report was – if only Facebook had been doing our job here.”
It is the usual problem when people who do not understand the world of espionage try to regulate it. The underlying issue seems to be that the thinking of these law makers is something like: “The intelligence services are very, very good. If they are not working properly, then the problem must lie elsewhere.” The fallacy with this is that the British intelligence and security services are not very good. In fact their record is extremely poor in comparison to other services around the world. This is hardly a secret. As an intelligence services historian I am frequently struck by just how much information about this there is in the public record. And yet the old myths persist and it is that, more than anything else I think, which prevents a proper examination of how to improve the efficiency of our intelligence system.
Why do those myths persist? A subject for a separate post… possibly.