The BBC says: "It is part of a plan to store details of all phone calls, e-mails and websites visited on a central database," adding that "Civil liberties campaigners have called the proposals a 'snoopers' charter'," which is predictable enough.

It's true, of course, that such a system would enable the authorities to snoop on everything we do, and any safeguards put into effect would likely not stand up to the holes certain to be present in the system (the government has a lousy track record on big IT projects - just look at the NHS National Programme for IT - and it has form in this area).

Now, the usual response from the other side of the fence is: "If you've nothing to hide, you've nothing to fear". This is fatuous and disingenuous. We talk and joke over email and the web about things we'd never do - there are plenty of George Bush jokes online, none of them written by people who want to kill the former American president.

Besides, as the comedian Robert Newman points out, the logical conclusion of that argument is that anyone who has anything to hide should be persecuted.

But then even if the proposals get put into place, there's the 'little elves' problem to worry about. This was theorised by John Grisham, of all people, in his novel The Brethren:

There were two types of phones at [the prison] Trumble, secured and unsecured. In theory, all calls made on unsecured lines were taped and subject to review by little elves in a booth somewhere who did nothing but listen to a million hours of useless chatter. In reality, about half the calls were actually taped, at random, and only about 5 percent were ever heard by anybody working for the prison. Not even the federal government could hire enough elves to handle all the listening.

A real-life example of this can be found in Peter Wright's controversial 1980s docu-drama Spycatcher:

So much raw intelligence was flowing out from the East that it was literally swamping the resources available to transcribe and analyse it. MI6 had a special transcription center set up in Earl's Court, but they were still transcribing material seven years later when they discovered that George Blake had betrayed the Tunnel to the Russians from the outset.

Both of the above quotes come from Quentin Campbell of the UK Crypto mailing list.

Aside from the question of how it'll be able to function (and how much that will cost), can it actually do its job?

The technique of going through this kind of information automatically is called 'data mining', and it can be used in quite interesting ways. The diligent Ben Goldacre has produced a fine analysis of why it can't work in this case:

Even with these infeasibly accurate imaginary tests, when you screen a general population as proposed, it is hard to imagine a point where the false positives are usefully low, and the true positives are not missed. And our imaginary test really was ridiculously good: it's a very difficult job to identify suspects, just from slightly abnormal patterns in the normal things that everybody does.

He points us in the direction of security guru Bruce Schneier's 2005 essay on the same topic, in which he says: "In hindsight, it was really easy to connect the 9/11 dots and point to the warning signs, but it's much harder before the fact. Certainly, many terrorist plots share common warning signs, but each is unique, as well."

So is there anything to worry about over this plan? Yes. The government's record on such projects makes it highly unlikely that the project will be useful for what it's designed, and highly likely that it'll be costly and will be open to abuse in specific cases. But the likelihood is remote that any of us will find ourselves in the British-American prison on Diego Garcia with unspeakable things being done to our nether regions, simply because in this case probability is on our side, not theirs.

Picture from Flickr user André-Pierre.