How does the Dutch public feel about automated decision-making?
Artificial Intelligence (AI) opens the door to automated decision-making (ADM). This development has given rise to both hopes and concerns regarding the impact that ADM might have on our society, and hence to a call to investigate this impact. A national research team headed by the University of Amsterdam has contributed to this investigation by conducting a survey into the Dutch public's knowledge and perceptions of ADM and AI.
In order to gauge the public's knowledge, perceptions, hopes and concerns regarding AI, the research team administered a survey to a high-specification sample of the Dutch population. The results of this survey can be summarised as follows:
- The majority of respondents know little about AI and algorithms.
- The main concerns regarding AI pertain to manipulation, risks or unacceptable outcomes.
- There are hopes for ADM in terms of its potential usefulness and its potential for the objective treatment of citizens (e.g. those seeking to take out insurance policies or mortgages).
- Health care, criminal justice, trade, the media and politics were mentioned as the main areas for possible implementation when considering AI-led ADM.
- Human oversight, human dignity, fairness and accuracy were mentioned as key values when considering AI-led ADM.
With this survey, the team has made a contribution to the public and political debate regarding AI and created input for further research into the fairness of automated decision-making.
Survey background
The survey and its report are the result of a research collaboration between the University of Amsterdam, Tilburg University, Eindhoven University of Technology, Utrecht University and Radboud University. The report is based on the answers of 958 respondents to a survey administered between 26 June and 4 July 2018.