to answer
the threat of a diner presenting somebody else’s card. So we see that a battery
of known answers to potential problems have been embedded into the scripted
play.
However, in the
world of security, there is always a new threat. We will now look at how scripted
plays are modified and evolved as security improves. As we just say, stealing a
chip card without stealing the personal identification number is of limited
value. Therefore, elaborate schemes have been developed to uncover the personal
identification number of a card, and actually, this has been so much of a
threat that it is already guarded against in some respect in the way we will
now explain. Actually, when the diner enters her personal identification number
in the restaurant banking terminal, the entry is very different from entering a
personal identification number in a regular computer, as opposed to a secure
core device. In a regular computer, a person enters a personal identification
number at the keyboard, the computer reads the personal identification number
from there, and then sends it wherever is needed. However, this creates a
security hole. The computer can decide to store the personal identification
number somewhere before sending it, and then, it can reuse it at will later on.
To guard against this, certified banking terminals have a personal
identification number entry mechanism that does not transit by the terminal
before being presented to the card. There is actually a direct physical link
between the personal identification number entry and the card, so that only the
card can see the personal identification number. Entry of the personal
identification number is therefore protected from onlookers, and yet another
threat has been answered and is included in the scripted play.
But now, we’ll
look at the next step the fraudster may take, and what we’ll describe has
actually occurred, and is still occurring today. By tampering with the
terminal, it’s possible to steal the personal identification number. This has
been done in various ways, from the simple idea of spying the personal
identification number entry with, say, a tiny camera, to elaborate ways
involving superimposing the keyboard with a fake one, and even more pernicious
ways consisting in designing a bogus terminal embedding the first one. We will
not detail those here; suffice it to say that the personal identification
number is secure only to a point in attaching a card to a person. When we’re
talking about paying the restaurant, the fraud might be tolerable. When we’re
talking about using similar technology in higher security situations, we may be
talking about threats that cannot be accepted. That’s why the next idea has
been to include a fingerprint sensor in the card itself. As we described it
earlier, a tiny sensor is used either as an array where the finger is placed,
or a bar against which the finger is scanned. To come back to our restaurant,
with a fingerprint sensor on the card the banking terminal would never be
involved in the capture of the link between the card and the cardholder, and
that link would therefore be much more trustworthy than the personal
identification number; actually, in situations of higher security, the two can
be combined. With this new measure then, yet another set of threats are
addressed, and the scripted play can change.
Using the
traditional form of theater inherited from the Greeks as a metaphor for
interactions in a social ecosystem, we see that the competence of the system
(the script in this case) is fixed and predetermined, and the performance (the
actual play) maps very closely the competence. This model of interaction
defines a set way to answer possible threats, and is capable to guide us in
many of our activities. It is very close to the way computers know how to do
things today: instruct them in advance of all threats and answers, and they’ll
faithfully perform the play, albeit with different levels of performance
depending on their other strengths and capabilities, like their inherent speed
or communication bandwidth.
|