The role of artificial intelligence, or AI, is becoming increasingly important in healthcare. Yet for many, it is an unknown or intangible concept. What ethical risks are actually associated with the use of AI and how should we deal with it as a sector? The European Institute of Technology & Health (EIT Health) recently published a pilot study on this. Paul Timmers, Chief Advisor of EIT Health and former director of the European Commission, explains, “These are challenging but essential questions.”
What kind of study is it?
“This is a progressive and significant study because it shows how innovators in EIT Health view European guidelines on AI and ethics. The guidelines were published last April by the European Commission’s High-Level Expert Group on AI. Among other things, they deal with AI challenges such as respect for individual autonomy, data protection, transparency of AI algorithms, contribution to well-being, and accountability. The guidelines are general in nature and not written specifically for the healthcare sector. This research reveals for the first time, if they resonate (or not) with innovators in healthcare.”
- AI and Ethics in the Health Innovation Community, EIT Health
- Ethics Guidelines for Trustworthy AI, High-Level Group, European Commission
- Op-ed on AI Regulation and Innovation, Paul Timmers
What is Vilans’ involvement in this?
“The study consists of three parts: the responses from some 80 startups and other innovators to an extensive questionnaire, a series of case studies, and an analysis of both, with a number of recommendations. Within Vilans, Dr. Henk Herman Nap and Dirk Lukkien have taken on the case studies and carried out essential scientific analysis and recommendations.”
And what has been the result?
“First, only 22 percent of respondents had heard of the guidelines, although that is not surprising because they were only recently published at the time of the study. About 60 percent expect that some form of AI legislation will ultimately apply to their AI innovation. The highest priority is given to privacy and data management, security, traceability, and individual autonomy. In general, the ethics of diversity were considered less important and it was assumed that existing accountability procedures are adequate. A second interesting finding is the fact that many of the respondents are already very active with AI and ethics, and present practical examples. These can be adapted and shared (one of the recommendations). Finally, the research also highlights dilemmas such as choosing between the interests of the individual and the general public.”
The highest priority is given to privacy and data management, security, traceability and individual autonomy.
What can we conclude from this study?
“If we look at this study from a distance, it is striking how much practical information can be extracted even from such preliminary research. Innovation in health and healthy aging can provide great insight for a nuanced and flexible approach to combining AI regulation with innovation. The goal is – together with healthcare practitioners and innovators – to make it a reality. How do practitioners take AI ethics into account, how can others learn from these best practices and what are the areas for improvement? The hope is that we can now go one step further in this research to deepen our insights into the challenging but essential questions about AI and innovation, and that Vilans will once again make an essential contribution to this.’
The role of Vilans
Earlier Vilans announced that in the coming years it will focus on informing the sector about developments, research and implementation projects, and applications of AI in long-term care. Such (follow-up) studies will also support the sector in examining and choosing a meaningful and responsible direction for AI. The project, ‘Data-driven Care’ also focuses on the role of data and ‘smart’ technology in long-term care, for example in the article on AI Applications in the Electronic Client File. Articles will soon be published in which the increasing role of AI in long-term care is described in more detail, and in which various challenges and dilemmas surrounding AI are highlighted.