Initial Overleaf Import

whitepaper.tex

@@ -0,0 +1,216 @@
+%%% LaTeX Template: Two column article
+%%%
+%%% Source: http://www.howtotex.com/
+%%% Feel free to distribute this template, but please keep to referal to http://www.howtotex.com/ here.
+%%% Date: February 2011
+
+%%% Preamble
+\documentclass[	DIV=calc,%
+							paper=a4,%
+							fontsize=11pt,%
+							twocolumn, draft]{scrartcl}	 					% KOMA-article class
+
+\usepackage{lipsum}		
+
+
+
+\usepackage[english]{babel}										% English language/hyphenation
+\usepackage[protrusion=true,expansion=true]{microtype}				% Better typography
+\usepackage{amsmath,amsfonts,amsthm}					% Math packages
+\usepackage[pdftex]{graphicx}									% Enable pdflatex
+\usepackage[svgnames]{xcolor}									% Enabling colors by their 'svgnames'
+\usepackage[hang, small,labelfont=bf,up,textfont=it,up]{caption}	% Custom captions under/above floats
+\usepackage{epstopdf}												% Converts .eps to .pdf
+\usepackage{subfig}													% Subfigures
+\usepackage{booktabs}												% Nicer tables
+\usepackage{fix-cm}													% Custom fontsizes
+
+\usepackage{draftwatermark}
+\SetWatermarkText{Draft}
+\SetWatermarkScale{10}% Package to create dummy text
+
+%%% Custom sectioning (sectsty package)
+\usepackage{sectsty}													% Custom sectioning (see below)
+\allsectionsfont{%															% Change font of al section commands
+	\usefont{OT1}{phv}{b}{n}%										% bch-b-n: CharterBT-Bold font
+	}
+
+\sectionfont{%																% Change font of \section command
+	\usefont{OT1}{phv}{b}{n}%										% bch-b-n: CharterBT-Bold font
+	}
+
+
+
+%%% Headers and footers
+\usepackage{fancyhdr}												% Needed to define custom headers/footers
+	\pagestyle{fancy}														% Enabling the custom headers/footers
+\usepackage{lastpage}	
+
+% Header (empty)
+\lhead{}
+\chead{}
+\rhead{}
+% Footer (you may change this to your own needs)
+\lfoot{\footnotesize \texttt{HowToTeX.com} \textbullet ~Two column article template}
+\cfoot{}
+\rfoot{\footnotesize page \thepage\ of \pageref{LastPage}}	% "Page 1 of 2"
+\renewcommand{\headrulewidth}{0.0pt}
+\renewcommand{\footrulewidth}{0.4pt}
+
+
+
+%%% Creating an initial of the very first character of the content
+\usepackage{lettrine}
+\newcommand{\initial}[1]{%
+     \lettrine[lines=3,lhang=0.3,nindent=0em]{
+     				\color{DarkGoldenrod}
+     				{\textsf{#1}}}{}}
+
+
+
+%%% Title, author and date metadata
+\usepackage{titling}															% For custom titles
+
+\newcommand{\HorRule}{\color{DarkGoldenrod}%			% Creating a horizontal rule
+									  	\rule{\linewidth}{1pt}%
+										}
+%%begin novalidate
+\pretitle{\vspace{-30pt} \begin{flushleft} \HorRule 
+				\fontsize{50}{50} \usefont{OT1}{phv}{b}{n} \color{DarkRed} \selectfont 
+				}
+\title{TOWARDS AN ETHICS-BY-DESIGN APPROACH IN DATA EXPERIMENTATION PROJECTS}					% Title of your article goes here
+\posttitle{\par\end{flushleft}\vskip 0.5em}
+
+\preauthor{\begin{flushleft}
+					\large \lineskip 0.5em \usefont{OT1}{phv}{b}{sl} \color{DarkRed}}
+\author{Till Riedel, }											% Author name goes here
+\postauthor{\footnotesize \usefont{OT1}{phv}{m}{sl} \color{Black} 
+					Karlsruhe Instititute of Technology (KIT)							% Institution of author
+					\par\end{flushleft}\HorRule}
+%%end novalidate
+\date{}																				% No date
+
+
+
+%%% Begin document
+\begin{document}
+\maketitle
+\thispagestyle{fancy} 			% Enabling the custom headers/footers for the first page 
+% The first character should be within \initial{}
+\initial{H}\textbf{ere is some sample text to show the initial in the introductory paragraph of this template article.}
+\section{Background}\label{background}
+
+Within the Horizon 2020 a number of data experimentation projects were
+set up to show the case and accelerate data innovation. We ourselves
+were funded by the EUHubs4Data project which consisted of 42
+``experiments''. Those experiments were led by innovative SMEs that were
+independently selected in the so-called `open calls'. They were
+supported by project members (so-called data innovation hubs or i-Spaces
+in our case) that were directly funded to provide an infrastructure for
+experimentation. Typically, this setting emulates a market situation
+using a public offering. However, being carried out within a research
+and innovation project, the situation differs because SMEs can use
+public funding to cover both covering cost of the offered and their own
+work.
+
+One of the core distinguishing aspects of the upcoming European data
+economy will be data and AI ethics. Consequently, all initiatives have
+put ethics on their agenda. So has ours. After three years and having
+monitored 42 very different projects all around data, it is time to
+review and reflect on our learnings.
+
+\section{Learning ethics for real}\label{learning-ethics-for-real}
+
+Ideally, we would be able to take our learnings straight from the lab to
+reality. As already touched upon, our project has been set up to emulate
+a market-oriented data economy; however, if you particularly look at
+ethics, we see great differences.
+
+One of the core pillars of ethics is understanding and accepting
+responsibility. In a cascade-funded setting this is not easy. As money
+is passed from the European commission to coordinator onto subcontracted
+SMEs, a legal and contractional regime is established that sets the
+playing field for many things that follow. Ethical actions require the
+choice to do things in the right way. This goes both for the SME
+performing the data experiment (which had to work with the services and
+data provided by the framework project to funding) and the data
+innovation hubs (which in turn had to work for the projects that were
+selected by external reviewers). In such a setting, there is not that
+much room for deciding for shared values (if you do not consider public
+money as shared value in itself).
+
+We think it is important to acknowledge that precompetitive and funded
+environments have a dynamic of their own. They are important for
+progress towards a true European data economy; thus, we will focus our
+analysis mostly on this setting. We do this in the hope that other
+project that have similar mechanics can learn from us. Constant progress
+also in developing ethical frameworks can also have a positive impact on
+the market.
+
+\subsection{Competition of values}\label{competition-of-values}
+
+\textbf{Funded projects must be governed by choices and finding partners
+that share values. Also, in a funded setting, we need positive
+competition around trustworthy and responsible data innovations.}
+
+To be clear: we have seen external reviewers choose, let us say,
+challenging projects. Often generating a major positive impact has the
+risk that if done wrongly, it may a trigger also negative impact. In our
+public report on the findings from the first open call, we have listed
+many different ethical challenges that we encountered. We did not feel
+prepared for all these challenges (from conducting medical trials to
+dealing with financial transaction data) and had to rely on the
+competence of the SMEs conducting the experiments. In many cases that
+worked out well, however, in many cases elimination of ethical risk was
+not an option because the SME was funded because it promised an output,
+and the data innovation hubs were funded to support the SMEs. Failure on
+both sides was not really an option in order to bring the overall
+project to an end after the fair and independent selection of the open
+call was finished. We learnt by putting more and more `terms and
+conditions'' online. However, a balancing of risk and impact during the
+selection process based also on the non-formalised values of the
+infrastructure would have (e.g. by sometimes selecting the second-most
+impactful experiment, if it has better compatibility with the
+self-determined competences and values of the infrastructure providers).
+
+\subsection{Compliance as a baseline}\label{compliance-as-a-baseline}
+
+\textbf{We cannot argue that legal compliance is a given, and ethics
+should only go beyond this.}
+
+Most of the risks detected by the ethics monitoring group of our project
+centred on GDPR compliance. With the GDPR in place for more than 10
+years, it would expect that it should be at the core of all data
+processing that involves data related to human data subjects. The
+reality is different. Even a common understanding of terminology is
+complex. If you look at anonymity, there are on one hand certainly
+difficult edge cases that have required more recent court rulings. An
+example might be the definition of personal data as clarified in the
+famous Patric Breyer case (ECLI:EU:C:2016:779). However, the reality is
+that the difference between anonymity and pseudonymity is often
+understood even in simple cases. Many big companies have invested in
+compliance, also due to clear requirements and considerable possible
+fines. The SME space, but also the research system, is from our
+experience only very slowly catching up given exemptions and lack of
+enforcement in this domain. This fact makes it very hard to establish
+ethics monitoring, saying repeatedly to people `I'm not a lawyer, but I
+think what you're doing might not be legal this way.' The argument that
+data coming from the EU is more `ethical' due to a clear legal regime
+does not hold. We have seen in our experiments multiple public data sets
+which were published by EU projects for which we could not clearly
+determine a legal basis.
+
+\subsection{Setting values}\label{setting-values}
+
+As said above, over the course of our project, the terms and conditions
+of the open calls evolved. Also, some service providers set conditions
+on what they would support and what not. Often those terms were only
+there to have a better lever at enforcing compliance, particularly with
+the contractual requirements of the grant (namely the ethical impact
+assessment of the project and evidence collection required by funding
+programme).
+
+\ldots{}
+
+
+\end{document}