..there can never be perfect security

Noticed this, an extract from a contribution on Comment is Free, by 'peitha'

'...there can never be perfect security.

Tell me, would you accept more deaths on London Tube trains or more surveillance etc to stop them (or some of them anyway)?

The classic mistake that mass surveillance is effective in stopping such attacks.

Take the example of a Tube (suicide) bomber.

When, exactly, are you going to stop him?

On the Tube train itself a la de Menezes? Too late; if he's smart the bomb explodes anyway (though I'm not going to explain why on a public forum - just accept that even an immediately killing head shot can be too late, OK?). Plus of course the risk of bystanders being shot accidentally as well if you do try a shot there.

OK, on the possibly crowded platform? Assume it's crowded because no suicide bomber is going to blow up an empty train late at night, he's invariably going to attack at rush hour, as the 7/7 bombers did, both because it garners more victims and creates more disruption so it's reasonable to assume the platform will be crowded. Too late for the same reasons. Ditto the entrance hall - remember the King's Cross fire? They tend to be crowded as well.

OK, at the entrance to the Tube Station? Let's put airport style security gates there, and create a nice large crowd waiting to go through - or, in the terminology of the terrorist, a nice large target. Even better - a target that was created by the action of the authorities themselves.

Remember also, if the bomber is in a crowd already, or moving through a crowd it can be extremely difficult to get a stopper shot except at extreme close range - which could easily become a kamikaze response from the pov of the armed policeman taking the shot when the bomb explodes. You volunteering for the job?

Hmmm, guess we have to stop him on his way to the Tube Station? So how did he get there? On a busy bus? 7/7 again. That's no good either. At the bus stop? Nope, no good.

OK, so we have to stop him before he even sets out on his attack to be certain to stop him. But to do that we have to already know who he is and where he lives.

So how does mass surveillance help with that?

That's not a mass surveillance problem, that's a standard humint and police problem. Mass surveillance, far from helping, may actually hinder that because of a likely distraction of resources into investigating the huge number of false positives it will throw up.

So how does mass surveillance actually help prevent suicide bombings? All it is actually good for is tracking back once the bomber has already identified himself by detonating his device as we saw with 7/7. hey were being filmes almost the whole way and didn't care. Why? Because they didn't care about being caught after the event! That kind of goes with the territory for a suicide bomber.

Assuming any reasonably competent cell structure by the terrorists, that gets us back to a standard humint and police problem - tracing known contacts etc, but only after the bombing.

Mass surveillance doesn't help prevent attacks because it is (a) reactive not proactive and (b) the resources needed to track everyone even before they are known to be a 'suspect' in a future bombing is simply impossible in the real world.

Even if by data-mining you can scan all surveillance footage the number of false positives generated will overwhelm you. Remember one of the 7/7 bombers had already been picked up as a 'person of interest' by intelligence but because he was deemed a low-level target he was not put under surveillance because they lacked resources to cast so wide a net.

So why not spend the money improving those resources not spending billions on mass surveillance which is easily defeated and only reactive?

So why don't you now try to explain how mass surveillance will help, rather than just assuming that it will and expressing that view abusively. Think it through and explain your reasoning'

See also Ben Goldacre, the Guardian's celebrated Bad Science columnist - 'Spying on 60 Million People doesn't add up.

You may also be interested in this story of infiltrating a violent extremist group