US, China Face Mutually Assured Destruction In Cyberwar

chicksdaddy writes with a tidbit from the RSA conference. From the article: "A panel of security and policy experts speaking at the RSA Conference in San Francisco on Wednesday said that, despite dire warnings about the information warfare capabilities of China and other developing nations, the risk of an all-out cyberwar is remote, and that the U.S. still holds many of the cards. Rather than trying to deliver a knock-out cyberwar capability, the U.S. should embrace the Cold War notions of containment and mutually assured destruction with advanced nations like China and Russia. Tried and true methods to win security from cyberattacks include international diplomacy, multilateral agreements that clarify the parameters for peaceful and hostile cyberactions and — of course — a strong offensive capability."

Turn off the pipes

[Skapare]

If an attack does come from overseas, just turn off the pipes (power off whatever devices the physical undersea fibers connect to). The problem with this is that they are likely setting up a massive botnet within the target country that cannot be blocked by such a method. And we do see that the government tends to not care to shut down botnets [slashdot.org], even now.

So what does MAD mean in this context?

[roc97007]

Trivially, an attack can be stopped or at least contained by simply pulling the network plug. I can't envision a scenario where we'd "lose the internet". We might lose connectivity to some areas for awhile, but it's not like there's some timed self-destruct code buried in Cisco firmware that could be activated en-masse.

...Is there?

They are the reason you are losing the "cyberwar"

[Hentes]

Idiots like these are the main reason Americans are so vulnerable to online threats.
I have written [slashdot.org] about why cyberwar is a false analogy, so mostly I will just repeat myself:

• There is no warfare, it's just a new method of espionage.
• There is no mutually assured destruction: cybersabotage is anonymous, thus you can't counterattack, and even if you could, an all-out attack would still not be enough for complete destruction. Cyberespionage is a slow game, to seriously disrupt a target infrastructure you would have to research it for years. And while you theoretically could try to collect vulnerabilities and then exploit them all at the same time in a single strike, it's not really feasible as systems get upgrade from time to time, and you collection would get obsolete after a while. In this conflict you have to grab an opportunity when you have one, a single devastating strike isn't practical.
• Also, destruction in this sense is a huge exaggeration, you can't do serious (compared to a real war) harm from the internet.
• The source of an attack doesn't have to be a nation, it can be anyone with an internet connection, which combined with anonymity makes diplomacy worthless.
• The only true method that works is to secure your fucking systems, and run regular whitehat tests (or, in their words, "cyberwargames") to identify possible vulnerabilities.