Where art thou, Eve? Experiences laying traps for Internet eavesdroppers

This paper describes a set of experiments we conducted to answer the question: just how prevalent is Internet interception? That is, if we sent our most sensitive information (bank information, passwords, etc.) in the clear, should we expect to regret it?

For a little over a year, we sent different types of Internet traffic over unencrypted channels between multiple clients and servers located at geographically diverse locations around the globe. Our messages contained seemingly sensitive and valuable information, including login credentials for banking sites, password reset links, etc. In total, we found no instances in which our information was acted upon by an eavesdropper.

This paper details the numerous challenges— technical, legal, and ethical—of setting up and maintaining a year-long, large-scale honeytrap. We discuss some fundamental limitations of such an experiment, and argue why our results should not be misinterpreted to suggest that message encryption is gratuitous.