PROTECT deliverable 1 (Internal) – The ethics of personalisation

Ana Fernandez Inguanzo, Paul Kuyer, Karen Vázquez, Rosalie Waelen , Michał Wieczorek

Research output: Book/ReportCommissioned reportResearch

Abstract

New technological developments of the 21st century, such as Artificial Intelligence (AI), Machine Learning (ML) or Human-Computer Interaction (HCI), mainly focus on the optimisation of personalized applications and on accurate algorithmic decisions and prediction models for individuals (Zanker et al. 2019). Personalisation means “a process that changes the functionality, interface, information content, or distinctiveness of a system to increase its personal relevance to an individual” (Blom 2000, p.1). Different areas that use this personalized technology are recommender systems such as Netflix, Youtube or Spotify; web personalization such as web analytics (Adobe, Google optimizer…) or Customer-Relation Management systems (Salesforce, HubSpot…); information filtering that are for example employed to filter the Internet with algorithms, or even personal e-mail filters, and adaptive texts such as screen readers or keyboard filters (Idem). Recourse to personalization, however, is not new and has been constantly used to sell and make products. An example has been exposed in an e-consultancy website, where in the 18th century a customer will arrive to the store and scream “hat” then leave in the carriage, and the retail’s workmen will refer to the records of that customer of the hat size and style preference (McCaig 2013). According to the author, a customer-business relationship was always personal, however with the raise of chain shops and products, things started to get impersonal. The beginning of online shopping was similar, companies consider that all customers where the same. What does the digital revolution add to traditional personalization? It dramatically increases the level and scrutiny of personalization. With increasing competition and high consumer expectation, there are multiple digital strategies where personalization is often used (Fenech & Perkins 2015). 21st century developments have created important changes in our society (Boxever 2015), and with them the improvement and accuracy of personalization. Its progress can be traced through marketing practices, such as customer-relationship management (CRM) developed in the 1990s, where email marketing campaigns were established to gather emails of individuals to send personalized advertisements. Initially, this implied the use of personal data with only name recognition, but today personalization is used in emails and advertisements that can be sent based on the individual’s activity online; for example, personalized advertisements are now based on your shopping habits, and Netflix recommendations are based from what you have watched in the past. New analytic tools and technologies facilitate personalization to a great extent. The introduction of cloud computing in 2007 allowed businesses to gather and analyse data in a much more efficient and accurate way. Moreover, thanks to ML and AI, computers can process huge quantities of data within milliseconds, and are also being used in marketing to improve relevance of products and services to consumers, while digital assistants like Siri and Alexa are offering tailored information and recommendations in our homes, making personalization a household term and everyday occurrence. Consequently, the majority of businesses understand the usefulness of personalization and the need to personalize their products and campaigns. Different technologies are increasingly facilitating high levels of personalization, far from mere personalized advertisements, while many technologies and businesses include personalization in their designs. However, this is not exempt from criticism as the practice of personalization raises ethical and legal issues that must be analysed. Questions arise concerning the limitations of personalization practices and the protection of consumers against excessive, ubiquitous, or unwanted personalization. And even in cases where personalization is explicit, consensual and done as a good working practice, adverse effects can occur as the process is not always effective and can lead to mismatching services and products to users. It is, therefore, necessary to study the ethical and legal implications of personalization, in order to highlight and address the adverse effects it can have on our society. This work package consists of 12 case studies discussing different personalized technologies and their ethical and legal implications. Our goal is to emphasise the importance of personalization today and the impact it will have in the future. We will do so by exhibiting a diverse range of technologies that utilise personalization, showing how these technologies work and discussing which ethical and legal issues they raise.
Original languageEnglish
PublisherInternal for the European Commission
Number of pages126
Publication statusPublished - 1 Nov 2020

Fingerprint

Dive into the research topics of 'PROTECT deliverable 1 (Internal) – The ethics of personalisation'. Together they form a unique fingerprint.

Cite this