Proceedings of the 15th EICAR Annual Conference "Security in the Mobile and Networked World", Hamburg, Germany, 29 April - 2 May 2006 (Eds. Paul Turner and Vlasti Broucek), pp. 150-163. ISBN 87-987271-8-4.
Late draft (pdf version)
Abstract: Unsolicited commercial email or spam is recognized as a problem disrupting email communication and costing the community dearly. In order to protect recipients from receiving spam, anti-spam measures building on technologies, such as filters and block lists, have been deployed widely. There is some evidence that certain anti-spam measures based on the purported origin of the spam cause unintended consequences which relate to issues of equity of access which we term digital redlining. Spammers have an interest in bypassing such measures by obscuring the real origin of their messages. Investigating these effects means we need to determine the true origin of spam, despite the efforts of spammers to confuse us and spam filters. The aim to find the true origin of spam is different from the objective of most anti-spam developers who are mainly interested in identifying spam when it knocks on their front door (mail server). In this paper we discuss why the difference between originator and delivery host matters when investigating digital redlining. We also highlight some of the difficulties we are facing when trying to determine the originating host as opposed to the delivering host.revised 10 August 2006