Is the General Data Protection Regulation (GDPR) adequate in ensuring responsible innovation using data analytics? What is the role of ethics with regard to placing limits on technological developments? Does innovation drive business and industry transformations, or does shareholder value maximization drive innovation? These questions were raised at the valorization workshop. It is intended that insights gained from academic research are made available and will be valuable for economic or societal application at private and public institutions.
The use of data analytics regularly sparks debate; among other things on the privacy of citizens and the security of data. The aim of the SCALES research project was to support private and public partners in designing an institutional framework with appropriate checks and balances. With this goal in mind, researchers from Leiden Law School conducted case studies, providing academic insights and perspectives to the internal discussions of the partners concerning the application of data analytics and its societal impact.
At the valorization workshop on 22 August, these case studies were presented by the representatives of the project partners. Marleen Schippers presented IkDus – an initiative that she has worked on together with project partner, Artificial Intelligence company Target Holding. IkDus is an initiative to set up a personal digitalized health environment supporting ‘patient empowerment’, moving towards more efficient and cheaper healthcare. Ms Schippers showed the audience what challenges IkDus was facing as a data-sharing platform and described the limited role the GDPR (through application of the right to data portability) can have for incentivizing IkDus-like platforms.
Bas de Vrind, representing another project partner Alliander, presented a case study concerning ‘consumer empowerment’. Alliander is the largest Distribution System Operator (DSO) on the Dutch energy grid that is currently designing the customer consent management mechanism that would give energy customers the possibility to control data generated on their smart metre. At the end of the presentation, Mr De Vrind listed the regulatory challenges Alliander is facing during the implementation of such a consent management mechanism, including for instance presumed conflicts with the EU Energy Directive and right of access arising from the GDPR.
Jenneke Evers, PhD candidate at eLaw Center of Law and Digital Technologies, presented her ongoing case studies about semi-automated decision-making systems in the policing domain of the Netherlands. The governmental applications she has studied include for instance SyRI – a system which the government of the Netherlands uses with the aim of linking personal data to detect fraud with tax benefits, and the ProKid 23- algorithmic system in the juvenile criminal justice system focusing on the early identification, registration and referral of young people with a high probability of committing a crime in the future. Ms Evers is researching scientific legal and philosophical aspects of such data analytics, in particular how they fit in the framework of non-discrimination laws. This is particularly important, as databases can often contain bias from old decisions that might be reinforced if not adequately supervised.
The workshop proceeded from purely legal solutions to considering the regulatory challenges at hand. Ron Boelsma, representative of the Dutch Police, presented the solution of the Dutch law enforcement for achieving responsible innovation through applied ethics for data analytics. The Dutch Police and Leiden University, together with TU Delft Design for Values, have recently completed the research project publishing a white paper on “AI and Ethics at the Police: Towards Responsible Use of AI at the Dutch Police.” The importance of institutional ethics was highlighted by the guest speaker from IBM, Rob Nijman. Mr Nijman emphasized the importance of public trust and explained the long tradition of ensuring such trust with regard to the technological innovations at IBM by ensuring fairness, transparency, and accountability in their systems and organizational structure.
The workshop concluded with a panel discussion, where presenters discussed ways to operationalize high-level ethical principles in the everyday business models of their organizations. From the set of case studies we conducted with our partners, we concluded during the workshop that our partners’ initiatives do actively incorporate existing legal regulations and standards concerning data protection. At the same time, developing services and products based on data analytics, and opening up new markets for them, is proving to be challenging and is often still only in the early stages. We use regulation in the broad sense, so besides legal steering we also consider steering through market, architecture and social norms (i.e. ethics). For instance, during the workshop we observed that the social norm of trust is a crucial factor for responsible innovation with data analytics. Where data innovations often promise to reduce uncertainty by providing control and prediction, the mechanisms themselves entail uncertainty about their exact functioning and impact. Trust is a strategy to deal with uncertainty and is necessary for innovations to be successfully adopted. Nevertheless, for trust to be given it has to be earned – which creates further research and motivation for achieving more trustworthy applications of data analytics.