The Triumph of Profiling. Andreas Bernard

Чтение книги онлайн.

СКАЧАТЬ “very interesting parallels.” “While wrestling with questions concerning customers, acquisition, and marketing,” he goes on, “I was struck by an idea: New methods and options would be needed to understand the actual motives of consumers. It would have to become possible to enter the customers’ minds!”³⁹ The term “customer profiling,” however, is an explicit expression of what has become a fundamental business model in the digital age of search engines, social networks, and online services. As is well known, global firms such as Google and Facebook have built their empires on the promise, first formulated by Andrew Weinreich, that companies are able to tailor, with previously unthinkable flexibility, the form and frequency of their advertisements for individual users. In 2007, for instance, Google patented a controversial method that enabled the company to construct reliable “user profiles” from the behavior of computer game players. Along with a player's individual tactics, the amount of time he or she spends on a gaming platform supposedly provides insights into the user's consumer preferences and thus enables highly effective advertisements to be placed at specific points in time.⁴⁰

Over the past 15 years, the fact that companies possess such profile-based knowledge has provoked specific questions concerning data protection. In a country such as Germany, which has complex legal provisions governing “informational self-determination,” the development of the profile format was greeted with critical commentary from early on. As early as the year 2000, in an article on the use of personality profiles for marketing purposes, the lawyer Petra Wittig formulated grave “legal objections” against this form of data processing, and she stressed her conviction that even the consent of customers would do nothing to change the problematic nature of such a practice. The individual right to self-determination loses its validity, according to Wittig, “when personal integrity is placed without restriction at the disposal of those interested in data.” The creation of “user profiles” in marketing would therefore have to be forbidden in principle on the basis of the fundamental right to “informational self-determination.”⁴¹ Despite this early criticism, however, the precise legal status of the format was slow to be defined. As Christoph Schnabel noted in his 2009 dissertation on data protection and the concept of the profile, “there has yet to be a single case of legislation in which the creation of a profile has been treated as unlawful.” As of 2009, according to Schnabel, the format of the user profile did not even qualify as a “legal concept”⁴² – a loophole that was not closed until recently. In May, 2018, the European Union's long-discussed “General Data Protection Regulation,” which attempts to strike a balance between free trade and legal stability, finally came into effect. This regulation, which unifies the previously heterogeneous legislation of the individual member states, provides the terms “profile” and “profiling” with their first legal definition. They consist of “any form of automated processing of personal data evaluating the personal aspects relating to a natural person, in particular to analyze or predict aspects concerning the data subject's performance at work, economic situation, health, personal preferences or interests, reliability or behaviour, location or movements.” Among the “principles of fair and transparent processing” of personal data, EU regulations now “require that the data subject be informed of the existence of the processing operation and its purposes.”⁴³

The authors of this legislation were aware, however, that the knowledge in profiles is not simply being collected by companies but is also being produced by the individual users themselves. A preliminary remark thus states: “This Regulation does not apply to the processing of personal data by a natural person in the course of a purely personal or household activity and thus with no connection to a professional or commercial activity.” Such activity might include, for instance, “social networking.”⁴⁴ Aside from the fact that this stipulation maintains the fragile boundary between “personal” and “professional” activity online, the permeability of which has engendered many ways of economizing private life in digital culture, this passage does much to underscore the ambivalence of the current concept of the profile. For data protectionists, the superimposed “profile” is never fully congruent with the core of an individual's legally protected “personality.” Profiling is regarded as an external act of attribution, a fact that led Schnabel in 2009 to the conviction that, “as regards profiles, the self-determination of consumers is, from an economic perspective, diametrically opposed to the interests of businesses.”⁴⁵

      This constellation has since changed entirely. By way of their profiles, users of social media now endeavor on a daily basis to depict their own personality in a congruent manner, and in this act of self-determination they provide businesses and advertisers with a constant stream of information. Passive and active access to the format has yielded a remarkable alliance that can no longer be understood in terms of the traditional categories of data protection. In today's profiles on Facebook, LinkedIn, or Instagram, self-representation and external control – subjectivization and objectivization – are blending together in an inextricable manner, and we are only gradually beginning to see what new sorts of social and political spheres might arise from this alliance. Though less than 20 years old, in any case, Petra Wittig's suggestion that even the voluntary creation of profiles should be prohibited by law sounds like something from a distant era.

The American psychologist Michal Kosinski has assembled a wealth of evidence regarding the proximity of autonomous creation and external evaluation in today's profiles. Since 2011, he and his colleagues have published a number of articles concerned with making reliable statements about people by applying the methods of personality psychology to Twitter or Facebook profiles. Kosinski's analyses are based on the so-called “big-five” or “five-factor” model, which, since Lewis Goldberg's work in the late 1980s, has been a significant testing procedure in the field. The big-five model divides individual feelings and emotions into five basic traits and aims to determine, by means of a standardized set of questions, the relationship among these traits in the behavior of the person being tested. In this way, it hopes to construct a taxonomy of the human personality. Kosinski's much-discussed thesis is that such knowledge can be obtained far more quickly and with the same level of precision by analyzing profiles on social networks. In his first article, from 2011, he demonstrated with a small cohort of a few hundred users that the most important elements of a Twitter profile – its number of followers, the number of accounts followed by the user, and the number of tweets – were sufficient for determining someone's personality traits according to the five-factor model, and that the conclusions drawn in this manner corresponded to those determined in actual analyses with a probability greater than 90 percent. Twitter profiles, in other words, could be used to make reliable predictions about the personality types of the users in question – whether they are more or less “reserved,” “conscientious,” “agreeable,” “cooperative,” or “sensitive.”⁴⁶

In the following years, Kosinski and his colleagues expanded the scope of their investigation by inviting, on a Facebook page called “myPersonality,” hundreds of thousands of social media users to take a big-five personality test, and then they compared these results with the users’ profiles. In 2013, they published a study that analyzed the personality types of around 60,000 subjects in light of their “likes” on Facebook – that is, their affirmational responses to comments or to shared products, texts, pictures, and videos. “We show,” the authors claim, “that easily accessible digital records of behavior, Facebook Likes, can be used to automatically and accurately predict a range of highly sensitive personal attributes including: sexual orientation, ethnicity, religious and political views, … age, and gender.”⁴⁷ From the behavior discernible СКАЧАТЬ