Page 102 - Privacy in a Cyber Age Policy and Practice
P. 102

THE PRIVACY MERCHANTS  89

           medical or financial records becomes better protected online, less sensitive—
           and therefore, less protected—information can reveal volumes of sensitive
           information thanks to PVT. As Marcy Peek points out, “The Internet has
           allowed commercial decision-makers to manipulate technology in such a
           way as to identify persons according to a multitude of variables and catego-
               89
           ries.”  Each page Internet users visit and each ad they click on tracks their
           unique IP address to create a detailed portrait of their offline persona. Peek
           explains, “Through various means such as cookies, Web bugs, and personal
           data input such as zip codes, corporate marketers can obtain a person’s
           demographic and other information and ‘tag’ an individual on the basis of
           such information.” The individual is then categorized and ranked against
           other users. The result is “Weblining,” an online version of the offline dis-
           criminatory practice of “redlining” individuals by denying or increasing
           the cost of services based on their demographic. After the Fair Housing Act
           of 1968 prohibited redlining, which used a mortgage applicant’s neighbor-
           hood to discriminate along racial lines, banks used instead other markers
           of race as a basis for racial discriminations, relying on the social clubs people
           joined or churches they attended, for example. In other words, items of
           information that were not sensitive were used to divine other information
           that was meant to be private. Easy access to this type of nonsensitive infor-
           mation streamlines this practice.
             As early as 2000, Business Week highlighted a PVT service, offered by
           data broker company Acxiom, called InfoBase Ethnicity System, which
           matched names against information about housing, education, and income
                                                                    90
           in order to identify the unpublicized ethnicity of an individual or group.
           More recently, a computer consultant named Tom Owad wrote a simple
           piece of software allowing him to download public wish lists that Amazon
           customers post to catalog products they plan to buy. He downloaded
           over 250,000 wish lists in one day, used Yahoo’s People Search to identify
           addresses and phone numbers, and published a detailed map that showed
           the locations of people interested in certain books or themes. Owad
           explained, “It used to be you had to get a warrant to monitor a person or
           group of people. Today, it is increasingly easy to monitor ideas. And then
                                 91
           track them back to people.”  Most people who put simple items of infor-
           mation about their preferences on their Facebook profiles are unlikely to
           know that they can be used to divine their personality traits with 90 percent
           accuracy, as if they had taken personality tests. 92
             All of this suggests that laws that ban the use of sensitive information but
           do not require any action on the part of the millions of effected citizens—
           the way medical, financial, and select other records are now protected—
           could be reinforced by banning PVT of protected information. That is, the
           wall that separates more and less sensitive types of information could be
   97   98   99   100   101   102   103   104   105   106   107