I can only assume that the reasoning behind such an idea is that if the country is under attack by a foreign country who have infiltrated key government/military infrastructure and are actively stealing confidential information and the victims aren’t taking action (possibly because they have not realized the theft is occurring) a central government agency might be tasked with monitoring that no confidential data is being stolen, and therefore could hypothetically take swift action rather than having to contact the victims and mobilizing them to act. However, if this is indeed the case, it would still be an overkill. I would imagine that if they can detect such action, then they can block it too; no need for a wide net.

@Elbert Gelton – I agree. No one person / organization should have such power. The infrastructure should ultimately be able to defend itself. Personally I do not believe centralized intervention is needed. If some of the IT infrastructure is under attack it should be quite capable of defending itself even by cutting outside access if required. It should be the victim that controls the response to the attack however.

@Gary – I am not sure I can agree with you. You’re right in that it’s a defensive mechanism and not an offensive one or a retaliation, but I am not so sure that it prevents further damage. Obviously I don’t know when such a system is going to be developed (if ever) but unless it’s nationwide I doubt it will be effective. Disconnecting the US from the rest of the Internet provided it is indeed nationwide will, in my opinion, be more damaging then any attack. Furthermore it introduces a huge security risk in that anyone intent on disrupting IT operations in the US need only launch an attack to trick the victim into triggering the kill switch. In essence a clever attacker intended on disruption might achieve his goal by exploiting the country’s response to an attack rather than depending on the success of an attack, which in most cases is pretty limited.

I can understand the analogy, but I think comparing an internet kill switch to a nuclear warhead is a bit of an exaggeration. If you think about it, the purpose of an internet kill switch is to prevent further damage during a cyber attack on the nation’s online infrastructure. This manages to working functions intact while the attacker is kept at bay by other systems (which may or may not be physical). It isn’t an act of retaliation by any means. It’s simply a defensive mechanism.

It depends how widespread such a kill switch will affect obviously but to be really effective I would imagine such a kill switch would need to isolate all of America from the rest of world. In terms of destructiveness it would probably be more then a nuclear warhead. At least it’s environmentally better since flipping the switch back will at least get things back to normal without any harmful radiation lingering for years to come.

Am I the only one getting the feeling that it’s like having access to a nuclear warhead, but this time, the only viable targets are completely online?

To cut it short, I think these are all attacks that one should already be protecting against regardless of terrorism.