Data Leakage

PUBLISHED BY THE IEEE COMPUTER SOCIETY 1540-7993/04/$20.00 © 2004 IEEE IEEE SECURITY & PRIVACY 23 I t is well known among the more technical computer user communities that documents written and stored in the Microsoft Word document format might contain hidden data. Some communities of regular computer users are also aware of this fact—for example, people working in the legal sector. However, awareness of this problem is not sufficiently widespread, especially among nontechnical computer users. Such people typically use computers at work and at home but are not computer professionals; thus, they are probably not aware of the dangers that opaque proprietary file formats can carry in their hidden payloads. This article demonstrates how to automate and scale the search for this hidden text effect. The tools we use are simple and do not require specialized tuning or highly skilled input. We have (and assume) no knowledge of the Word document format itself, reinforcing the simplicity of the techniques discussed. We’ll also consider important but commonly neglected behavior patterns in the context of this problem.