🚀 go-pugleaf

In article <1992Sep8.115050.8694@cl.cam.ac.uk>, rja14@cl.cam.ac.uk (Ross Anderson) writes:

|> This sort of `false terminal attack' was first reported in the USA in
|> about 1988. You'd think the banks would warn people about it, but
|> instead they seem determined to introduce PIN pads in retail outlets 
|> here. The idiots seem to believe that this will cut fraud from the 
|> current \pounds 200 million plus, but in practice it means that every 
|> bent merchant in the country will be able to collect card and PIN data.

This type of attack has been going on for a lot longer than that, and it
doesn't require a fraudulent merchant to perpetrate.  There were a couple of
cases in the late 70s/early 80s of someone building a bogus ATM machine and
placing it directly in front of a real bank ATM machine.  I believe one case
was in New York and the other in Italy, but my memory may be flakey on this 
point.

This points up the need for one of the B2 and up security measures from
the orange book - namely trusted path.  There must be a way for the user
to authenticate to the system, but there must also be a way for the system
to authenticate itself to the user, so that user knows that this is the
real bank ATM or terminal or workstation or whatever...