Performing security verifications on a compromised system can give
a false sense of security. If compromised, a computer system can return
false results, thus "deceiving" the verification process. Our motivation for
this work is straightforward: Computers should not be trusted, at least
not when they are attesting their own integrity.
In our project Babel, this problem is addressed by, quite literally,
thinking outside the box. Babel introduces an architecture where the
user's computer is unable to execute any program by itself and depends on
an external entity to execute any application. Taking into consideration
the advances in computer network and cloud computing, we move the
verification process to outside the physical limits of the computer.
Babel can be mistaken for yet another instance of extant approaches.
In this paper, we revisit the Babel architecture with the twofold intention
of clarifying what Babel is and showing how Babel differs from previous