Question response at Harvard

Was anyone else surprised at the answer given to the question I asked
at the class in Harvard? My question addressed the technical
feasibility of the encryption backdoor, more specifically, the concept
of a key bank. I claimed that creating a hackproof system to protect
those keys would be impossible since no code can be perfect on the
first try.

The answer I received can be labeled as "The Club Hypothesis", where
although The Club is not unbreakable, it assures a car protection over
a car without a Club which is, consequently, easier to break into. The
connection seemed to be that although the code protecting the bank
would not be unhackable, it would be advanced enough to instead
encourage attacks on other systems which would be less protected.

Unfortunately, since time was running out, I could not ask Mr. Zitrain
to elaborate on his answer. I must admit that I was disappointed with
his explanation, though. Isn't one of the principles of hacking the
challenges it offers, as opposed to the ratio of time and rewards?

If this is the explanation protecting the feasibility of a keybank,
doesn't it assure an imminent disaster if it is ever implemented?

-- Cynthia Johanson, October 3, 1999

Answers

The club hypothesis

The club hypothesis doesn't apply here. It assumes that all cars are of more or less equal value, and so the deciding factor of which car to break into will be the one without the club. However, in the case of a large centralized keybank, the value of attacking the keybank is much larger than the value of attacking any single system which is protected by a key in the bank. Sure, if I want to steal credit card numbers, it might be easier to get someone's number by social engineering, burglary, or otherwise than it would be to break in to the the keybank. However, at the end of the day what do I have? One lousy credit card number.

If I break into the keybank, although difficult as it may be, what have i got? Every creditcard number. Maybe even every social sec number, every bank number and PIN number. Maybe a national economy. What's that worth? A little more than a one lousy credit card number?

I think so.

-- Aaron VanDevender, October 5, 1999