Ethics of Public Use of AI and Big Data

The Case of Amsterdam’s Crowdedness Project

  • Mark Ryan University of Twente


Smart information systems (Big Data and artificial intelligence) are used by governments to improve mobility, reduce over-crowdedness in hotspots, and provide more effective management of crowds. I looked at how Amsterdam municipality is using smart information systems (SIS) in their DrukteRadar Project to identify, report, and tackle issues surrounding crowdedness levels in the city.

SIS are becoming popular amongst governmental officials to automate activities more effectively. SIS provide the opportunity to improve mobility, increase economic growth, reduce energy outputs, improve management decisions, respond to disasters quicker, and improve citizens’ quality of life. They offer governments the possibility of improving services, while reducing costs. The use and implementation of SIS is becoming widespread and governments are observing the benefits posed by SIS, particularly in relation to urban management.  

80% of Europe’s population will live in cities by 2020 and governments face a huge strain on resources and infrastructure. The use of SIS is being pioneered to help governments meet these needs and to provide a sustainable future for urban citizens. Ethical issues in this context can include that data may not be accurate, faithful or representative of a city and its citizens, which may cause bias, prejudice and harm to a population, by leading to unfair service provision. ICT companies’ involvement in governmental SIS projects may also lead to technological lock-in and dependency on corporations. Instantaneous and ubiquitous retrieval and analysis of data may infringe upon citizens’ privacy and may lead vulnerabilities of malicious hacking, stolen data and a city’s security.

To uncover if these issues correlate with the experience of those working in the field, I interviewed the Project Owner of Amsterdam’s DrukteRadar project (translated as crowdedness project). This project implements SIS to anticipate and prevent overcrowding in Amsterdam, and was created in response to growing pressures on the city’s amenities. The DrukteRadar Project collates a wide array of datasets to predict crowd levels and potential problem hotspots, visualised through a digital dashboard. The project aims to improve municipality management, provide help to tourists planning their trips, and assisting citizens’ navigation through the city.

Through my discussions with the Project Owner of the DrukteRadar, I uncovered two additional issues not found in the literature: access to SIS and data ownership. The DrukteRadar team is concerned about access to SIS to promote fairness, equality, and provision of services amongst citizens. It aims to make its dashboard user-friendly and available to as many people as possible to promote inclusion. Data ownership is a concern for the project – who owns the data and what can be done with it. The DrukteRadar Project ensures they have data sovereignty, so that they do not become technologically locked-in to relationships with private organisations.

The Project Owner was aware that inaccurate of data may lead to discriminatory recommendations and harmful consequences. The DrukteRadar Project tries to minimise their algorithmic inaccuracy through extensive monitoring; secure technical infrastructure; and stakeholder review sessions. Another interesting finding was identifying how projects ensure privacy protection of its citizens. The DrukteRadar ensures that data is not traceable to individuals and the use of datasets follow privacy-by-design principles. The project also has strong security protocols, cyber-security measures, anonymization techniques, and repeated vulnerability tests. Overall, my report was able to evaluate how ethical issues found within the SIS literature correlate to those identified, and tackled, in practice. as well as highlighting the two additional concerns not explicitly mentioned in the literature.

How to Cite
Ryan, M. (2018). Ethics of Public Use of AI and Big Data. ORBIT Journal, 2(1).