Furthermore, a determined adversary with a well-funded and well-prepared attack wouldn't just attack the financial services industry, they'd attack other key industries.

So for example, their are 9 major oil refineries in the UK. If they could all be put out of action with Stuxnet-style attacks (or maybe with precision cruise missiles), then 2 weeks later the UK runs out of fuel and supermarkets run out of food.

Or consider an attack on the internet infrastructure itself. Since we're envisaging an attack in the future, when TV, radio and telephones are just services that run on the internet, bringing down the net will destroy communications. And the end-pints -- computers in homes and offices -- will themselves be down.

Societies are only going to get more dependent on computers, and computers are only going to get more complicated (making logic bombs harder to find). The damage that a cyberspace attack could do, and the ease of pulling it off, will therefore both only increase.

One more thing about this... you can use about the same kind of explosives on buildings and bridges just varying amounts and methods of application.

Whereas screwing over the whole banking system of north USA like that requires the kind of cracking you described but against maybe a few hundred different targets, likely all running different systems and posing different challenges - not un-do-able - but ultimately to not that great avail. So you have a lot of effort actually pulling the attack off to even make a dent and then it is probably going to be not much more than a scary headline in the newspaper.

Banks and governments are not the internet or in the internet. They are offering their services ALSO over the internet but not exclusively. And the really important backups need to be kept for years and years, off site. I doubt it is as easy and efficient to launch "cyber war" on the USA and then just "cyber nuke them back to the stone age" which is something this article subtly suggests.

Also, how many power plants, reactors, flood gates and traffic systems are controllable over the internet like that? At least here in Europe I have honestly no idea but strongly doubt they work like that.

Of course, you could equally well just sow doubt with a few precise surgical strikes. If random customers lose all their money often enough, the rest of the customers will withdraw theirs to have it in physical assets.

Which is enough to cause tremendous damage. And a way more likely scenario.

My point was: you cannot simply just like that loose money without any trace of what happened... it's not like you can just "DELETE FROM accounts" and it is totally un-tracably gone.

I would trust more in the people who design our SWIFT and other transaction systems.

I wouldn't. After seeing such tremendous security accumen from CitiGroup, I wonder how you can blindly trust.

But the larger point is, even if you can mostly undo the damage, you have a much harder time restoring trust. The money doesn't need to be lost. If you create enough of a panic to trigger a run on banks, you've achieved your goal, too.

I mostly agree with your points, but "They are offering their services ALSO over the internet but not exclusively." is irrelevant. The attacks being discussed are not about the public-facing services but about their internal internet usage for interbank transactions and linking branches of the same bank.